Compare commits

...

316 Commits

Author SHA1 Message Date
Raj Maheshwari 64be2cb466 [Fix] Minor bug in parse_response of CodeActResponseParser (#2912) 2024-07-13 14:36:27 +00:00
மனோஜ்குமார் பழனிச்சாமி b2b6d2ac1e Fix: hostname in logging (#2914)
Co-authored-by: tobitege <tobitege@gmx.de>
2024-07-13 10:04:08 +00:00
Boxuan Li f249254ba4 Fix delegator LLM config when config is set from UI (#2913) 2024-07-13 11:31:30 +02:00
Xingyao Wang 7e68de746d arch: refractor eventstream into async (#2907)
* deprecating recall action

* fix integration tests

* fix integration tests

* refractor runtime to use async

* remove search memory

* rename .initialize to .ainit
2024-07-12 20:03:28 +00:00
Xingyao Wang e45ddeb2a2 arch: deprecating recall action and search_memory (#2900)
* deprecating recall action

* fix integration tests

* fix integration tests

* remove search memory
2024-07-12 19:23:21 +00:00
Boxuan Li 2b7c4e5571 Remove legacy dummy action from CI (#2903) 2024-07-13 02:27:13 +08:00
Boxuan Li ebbc0e6803 Integration testing: unset irrelevant env variables (#2902) 2024-07-12 22:12:37 +08:00
Xingyao Wang 96b5cb78fd [Arch] EventStreamRuntime supports browser (#2899)
* fix the case when source and tmp are not on the same device

* always build a dev box (with updated source code) for development purpose

* tail the log before removing the container

* move browse function

* support browser!
2024-07-11 22:32:12 +00:00
Xingyao Wang ced7499f8d fix Runtime import (#2897) 2024-07-11 21:51:57 +00:00
மனோஜ்குமார் பழனிச்சாமி 7cbf2d9f16 Doc: LM Studio guide (#2875) 2024-07-11 23:17:03 +02:00
Xingyao Wang e45d46c993 [Arch] Implement EventStream Runtime Client with Jupyter Support using Agnostic Sandbox (#2879)
* support loading a particular runtime class via config.runtime (default to server to not break things)

* move image agnostic util to shared runtime util

* move dependency

* include poetry.lock in sdist

* accept port as arg for client

* make client start server with specified port

* update image agnostic utility for eventstream runtime

* make client and runtime working with REST API

* rename execute_server

* add plugin to initialize stuff inside es-runtime;
cleanup runtime methods to delegate everything to container

* remove redundant ls -alh

* fix jupyter

* improve logging in agnostic sandbox

* improve logging of test function

* add read & edit

* update agnostic sandbox

* support setting work dir at start

* fix file read/write test

* fix unit test

* update tescase

* Fix unit test again

* fix unit test again again
2024-07-12 01:52:26 +08:00
மனோஜ்குமார் பழனிச்சாமி 43c3e904b7 Restore last mute setting (#2895) 2024-07-11 16:48:04 +00:00
dependabot[bot] ae2fbbf8af chore(deps): bump boto3 from 1.34.142 to 1.34.143 (#2893) 2024-07-12 00:12:23 +08:00
dependabot[bot] 29ed1d744d chore(deps-dev): bump chromadb from 0.5.3 to 0.5.4 (#2892) 2024-07-12 00:12:11 +08:00
dependabot[bot] 217eed9dab chore(deps): bump litellm from 1.41.14 to 1.41.15 (#2891) 2024-07-12 00:11:59 +08:00
மனோஜ்குமார் பழனிச்சாமி 6bef270526 Doc: Fix Azure Guide (#2894)
* Doc: Fix Azure Guide

* Update azureLLMs.md
2024-07-11 15:56:36 +00:00
Xingyao Wang 1b54800a29 [Agent] Improve edits by adding back edit_file_by_line (#2722)
* add replace-based block edit & preliminary test case fix

* further fix the insert behavior

* make edit only work on first occurence

* bump codeact version since we now use new edit agentskills

* update prompt for new agentskills

* update integration tests

* make run_infer.sh executable

* remove code block for edit_file

* update integration test for prompt changes

* default to not use hint for eval

* fix insert emptyfile bug

* throw value error when `to_replace` is empty

* make `_edit_or_insert_file` return string so we can try to fix some linter errors (best attempt)

* add todo

* update integration test

* fix sandbox test for this PR

* fix inserting with additional newline

* rename to edit_file_by_replace

* add back `edit_file_by_line`

* update prompt for new editing tool

* fix integration tests

* bump codeact version since there are more changes

* add back append file

* fix current line for append

* fix append unit tests

* change the location where we show edited line no to agent and fix tests

* update integration tests

* fix global window size affect by open_file bug

* fix global window size affect by open_file bug

* increase window size to 300

* add file beginning and ending marker to avoid looping

* expand the editor window to better display edit error for model

* refractor to breakdown edit to internal functions

* reduce window to 200

* move window to 100

* refractor to cleanup some logic into _calculate_window_bounds

* fix integration tests

* fix sandbox test on new prompt

* update demonstration with new changes

* fix integration

* initialize llm inside process_instance to circumvent "AttributeError: Can't pickle local object"

* update kwargs

* retry for internal server error

* fix max iteration

* override max iter from config

* fix integration tests

* remove edit file by line

* fix integration tests

* add instruction to avoid hanging

* Revert "add instruction to avoid hanging"

This reverts commit 06fd2c5938.

* handle content policy violation error

* fix integration tests

* fix typo in prompt - the window is 100

* update all integration tests

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
Co-authored-by: Xingyao Wang <xingyao@all-hands.dev>
2024-07-11 15:30:20 +00:00
dependabot[bot] e793ca2261 chore(deps): bump framer-motion from 11.3.0 to 11.3.2 in /frontend (#2890) 2024-07-11 23:19:52 +08:00
adragos 5f61885e44 feat: Implement user confirmation mode, request confirmation when running bash/python code in this mode (#2774)
* [feat] confirmation mode for bash actions

* feat: Add modal setting for Confirmation Mode

* fix: frontend tests for confirmation mode switch

* fix: add missing CONFIRMATION_MODE value in SettingsModal.test.tsx

* fix: update test to integrate new setting

* feat: Implement user confirmation for running bash/python code

* fix: don't display rejected actions

* fix: linting, rename/refactor based on feedback

* fix: add property only to commands, pass serialization tests

* fix: package-lock.json, lint test_action_serialization.py

* test: add is_confirmed to integration test outputs

---------

Co-authored-by: Mislav Balunovic <mislav.balunovic@gmail.com>
2024-07-11 14:57:21 +03:00
மனோஜ்குமார் பழனிச்சாமி 1d4f422638 Doc: Mention FORCE_REGENERATE var (#2833)
* Mention FORCE_REGENERATE var in doc

* Update tests/integration/README.md

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
2024-07-11 04:01:15 +00:00
dependabot[bot] 456690818c chore(deps): bump google-cloud-aiplatform from 1.58.0 to 1.59.0 (#2884)
Bumps [google-cloud-aiplatform](https://github.com/googleapis/python-aiplatform) from 1.58.0 to 1.59.0.
- [Release notes](https://github.com/googleapis/python-aiplatform/releases)
- [Changelog](https://github.com/googleapis/python-aiplatform/blob/main/CHANGELOG.md)
- [Commits](https://github.com/googleapis/python-aiplatform/compare/v1.58.0...v1.59.0)

---
updated-dependencies:
- dependency-name: google-cloud-aiplatform
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-10 10:09:48 -07:00
dependabot[bot] de11b49a38 chore(deps): bump boto3 from 1.34.141 to 1.34.142 (#2882) 2024-07-11 00:04:56 +08:00
dependabot[bot] 50f07aea44 chore(deps): bump framer-motion from 11.2.14 to 11.3.0 in /frontend (#2880) 2024-07-11 00:03:22 +08:00
dependabot[bot] 008f288bb0 chore(deps-dev): bump openai from 1.35.10 to 1.35.13 (#2885) 2024-07-11 00:02:55 +08:00
dependabot[bot] 7938b454e4 chore(deps): bump litellm from 1.41.13 to 1.41.14 (#2883) 2024-07-11 00:02:39 +08:00
dependabot[bot] 45e40d68f6 chore(deps): bump json-repair from 0.25.2 to 0.25.3 (#2881) 2024-07-11 00:02:17 +08:00
Boxuan Li c68478f470 Customize LLM config per agent (#2756)
Currently, OpenDevin uses a global singleton LLM config and a global singleton agent config. This PR allows customers to configure an LLM config for each agent. A hypothetically useful scenario is to use a cheaper LLM for repo exploration / code search, and a more powerful LLM to actually do the problem solving (CodeActAgent).

Partially solves #2075 (web GUI improvement is not the goal of this PR)
2024-07-09 22:05:54 -07:00
Jiayi Pan 23e2d01cf5 Fix instance agonistic: remove Miniforge after installation (#2878)
* remove Miniforge after installation

* fix typo
2024-07-09 20:08:58 +00:00
dependabot[bot] de47d8eecc chore(deps-dev): bump @typescript-eslint/eslint-plugin in /frontend (#2872) 2024-07-09 16:42:22 +00:00
dependabot[bot] 4049c69590 chore(deps): bump litellm from 1.41.11 to 1.41.13 (#2870) 2024-07-10 00:08:12 +08:00
dependabot[bot] 792949aeb2 chore(deps): bump boto3 from 1.34.140 to 1.34.141 (#2869) 2024-07-10 00:07:48 +08:00
dependabot[bot] 864ee465fe chore(deps): bump google-generativeai from 0.7.1 to 0.7.2 (#2868) 2024-07-10 00:07:35 +08:00
dependabot[bot] fbced43ff3 chore(deps): bump framer-motion from 11.2.13 to 11.2.14 in /frontend (#2871) 2024-07-10 00:07:12 +08:00
dependabot[bot] 46b853e1b4 chore(deps-dev): bump @typescript-eslint/parser in /frontend (#2873) 2024-07-10 00:06:36 +08:00
Yufan Song 9198ea3fb1 fix output code error in docker image (#2862) 2024-07-08 22:40:39 -07:00
Yufan Song f0bc231f3e chores: open the websockets ports for port mapping and remove chores. (#2864)
* add port map

* add more comments TODO
2024-07-08 22:33:58 -07:00
Yufan Song 8cfb1be5a3 od-runtime-client: check and remove permission TODO (#2863) 2024-07-08 22:33:25 -07:00
Xingyao Wang f2e92b2db7 move image agnostic util to shared runtime util (#2859) 2024-07-08 22:17:01 +00:00
mamoodi e2636f9ece docs: Reorder docs and small update to README (#2860)
Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
2024-07-08 21:33:12 +00:00
Yufan Song 351127db55 add od runtime clinet dependencies (#2858) 2024-07-08 18:20:44 +00:00
Yufan Song 9fbfa0650e Add websocket runtime and od-client-runtime (#2603)
* add draft code

* add some sandbox draft code

* Export WebSocketBox and fix add_to_env async

* fix

* test execute

* add runtime draft

* add draft od-runtime-client

* refactor useless code

* format

* resume runtime

* resume runtime

* remove background command

* remove uselss action and init function

* add EventStreamRuntime test

* add echo server test

* temporarily build websocket everytimes

* remove websocket sandbox deprecated

* refactor code

* fix bug, add test

* fix bug

* remove test draft code

* refactor code, remove async

* rename file and directory

* add init plugin and runtime tools function

* add docker luanch

* fix plugin initialization

* remove test scropt

* add mock test code

* apply suggestions

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-07-08 17:44:37 +00:00
dependabot[bot] e6ebb4307e chore(deps): bump boto3 from 1.34.139 to 1.34.140 (#2855)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.139 to 1.34.140.
- [Release notes](https://github.com/boto/boto3/releases)
- [Commits](https://github.com/boto/boto3/compare/1.34.139...1.34.140)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-08 15:43:42 +00:00
dependabot[bot] 9b9b754965 chore(deps): bump litellm from 1.41.7 to 1.41.11 (#2854)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.41.7 to 1.41.11.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.41.7...v1.41.11)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-08 15:41:41 +00:00
dependabot[bot] 648597c4f7 chore(deps-dev): bump @types/node from 20.14.9 to 20.14.10 in /frontend (#2852)
Bumps [@types/node](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/node) from 20.14.9 to 20.14.10.
- [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases)
- [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/node)

---
updated-dependencies:
- dependency-name: "@types/node"
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-08 14:49:49 +00:00
dependabot[bot] ff701f99e6 chore(deps): bump tailwind-merge from 2.3.0 to 2.4.0 in /frontend (#2851)
Bumps [tailwind-merge](https://github.com/dcastil/tailwind-merge) from 2.3.0 to 2.4.0.
- [Release notes](https://github.com/dcastil/tailwind-merge/releases)
- [Commits](https://github.com/dcastil/tailwind-merge/compare/v2.3.0...v2.4.0)

---
updated-dependencies:
- dependency-name: tailwind-merge
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-08 14:49:39 +00:00
Engel Nyst 2df1d67007 History clean up (#2849)
* clean up add_history

* refactor last agent message
2024-07-08 05:10:21 +02:00
மனோஜ்குமார் பழனிச்சாமி c6aa50779d Update regenerate.sh (#2832) 2024-07-07 23:52:03 +02:00
Ralf D. Müller ba0f57c279 added netcat to the requirements (#2822) 2024-07-07 21:32:56 +00:00
Engel Nyst d37b2973b2 Refactoring: event stream based agent history (#2709)
* add to event stream sync

* remove async from tests

* small logging spam fix

* remove swe agent

* arch refactoring: use history from the event stream

* refactor agents

* monologue agent

* ruff

* planner agent

* micro-agents

* refactor history in evaluations

* evals history refactoring

* adapt evals and tests

* unit testing stuck

* testing micro agents, event stream

* fix planner agent

* fix tests

* fix stuck after rename

* fix test

* small clean up

* fix merge

* fix merge issue

* fix integration tests

* Update agenthub/dummy_agent/agent.py

* fix tests

* rename more clearly; add todo; clean up
2024-07-07 21:04:23 +00:00
மனோஜ்குமார் பழனிச்சாமி 9dc2d2c80f Refactor: Remove extra log (#2687) 2024-07-08 05:37:13 +09:00
Shimada666 e35c1ff74a Display real-time build logs for the agnostic image (#2830)
* Display real-time build logs for the agnostic image and improve wget's output.

* remove unused code
2024-07-08 04:35:16 +08:00
மனோஜ்குமார் பழனிச்சாமி 34c765688b Streamline Logging Events (#2532)
* Skip duplicate log

* log user actions

* fix tests

* log all action _step

* refactor log

* revert test

* refactor log

* visual diff

* disable overriding event source

* Revert "disable overriding event source"

This reverts commit b0047cc0cd.

* Refactor logic

* refactored runtime on_event

* fix merge conflict

in Web UI, it shows as red color (seems deletion but added)

* linted

---------

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
2024-07-08 05:34:47 +09:00
மனோஜ்குமார் பழனிச்சாமி adf1a0d556 Bugfix: add missing f-string for logging debug message in task creation (#2836) 2024-07-07 17:36:19 +02:00
மனோஜ்குமார் பழனிச்சாமி 85a817304e Check exit code (#2834) 2024-07-07 17:35:22 +02:00
Graham Neubig d0384cafdd Two fixes to swe bench eval (#2831)
* Two fixes to swe bench eval

* Add error message

* Change dumping of metadata
2024-07-07 07:21:50 +00:00
மனோஜ்குமார் பழனிச்சாமி 3a3694ca17 doc: Mention negative feedback feature in bug report. (#2827)
* doc: Mention feedback feature in bug report.

* Update .github/ISSUE_TEMPLATE/bug_template.yml

Co-authored-by: Graham Neubig <neubig@gmail.com>

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-07-07 06:56:12 +00:00
Bin Lei c8e5848add fix git diff TIMEOUT problem in swe_bench evaluation (#2828)
* fix git diff TIMEOUT problem in swe_bench evaluation

* fix git diff TIMEOUT problem in swe_bench evaluation

* Update evaluation/swe_bench/swe_env_box.py

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>

---------

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>
2024-07-07 06:30:59 +00:00
Shimada666 0973e31f00 Update custom sandbox usage guide (#2829) 2024-07-07 05:33:35 +02:00
Shimada666 82f256be96 trim the sandbox image and install plugin dependencies in agnostic image (#2792)
* trim the sandbox image

* remove wrong code

* readd python

* readd python

* fix script

* readd nano
2024-07-06 17:38:37 +02:00
மனோஜ்குமார் பழனிச்சாமி d6570bd572 Fix gemini-1.5-flash crash due to missing 'vertexai' module (#2826)
* Fix gemini-1.5-flash crash due to missing 'vertexai' module

* Update poetry.lock
2024-07-06 16:27:59 +02:00
Shimada666 d22ff73905 Make the sandbox Python runtime completely transparent (#2796)
* Make the sandbox Python runtime completely independent

* fix source bashrc

* add pip install instruction for ipython to fix intergration tests for codeact swe

* update integration tests

* change flake8 command to (maybe) fix sandbox tests?

* make lint support both unittest & sandbox

* fix agnostic image build error

* refactor build script

---------

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
Co-authored-by: tobitege <tobitege@gmx.de>
2024-07-06 15:22:44 +02:00
Leo 9b0ff117ab CI: Support uploading frontend unit test coverage. (#2772)
* CI: Support uploading frontend unit test coverage.

* Add make-i18n before test.

* Update vitest configuration to include only .ts and .tsx files in coverage.

* remove .only in test and fix the failed tests.

* Add text summary.

* Move vite-tsconfig-paths to dev dep. Adjust UTs.

---------

Signed-off-by: ifuryst <ifuryst@gmail.com>
Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com>
2024-07-06 12:16:30 +08:00
Xingyao Wang f6dc89b41a [Evaluation] Simplify eval & and multi-processing related fixes (#2810)
* initialize agent inside process_instance_fn;

* remove dependency on `config.max_iterations`

* switch back to only include llm config to metadata
2024-07-06 07:18:46 +08:00
Xingyao Wang a47713ecb0 [Arch] Remove supports for Background Commands (#2803)
* depracting docker exec box

* remove doc exec from workflow and docs

* remove background commands

* Update tests/unit/test_sandbox.py

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>

* replace for-loop with assignment

* fix integration tests

* fix integration tests for shell script

* fix integration tests

* increase max iter to fix some monologue agent issue

* fix integration test again

* fix integration tests (seems related to run_user issue)

---------

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
2024-07-06 03:38:05 +08:00
mamoodi 99284da476 Use README as the only place for getting started instructions (#2815)
Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
2024-07-05 19:44:29 +02:00
mamoodi 9ccc64fa7e Update README (#2814)
Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
2024-07-05 12:54:34 -04:00
dependabot[bot] eda582335a chore(deps): bump tenacity from 8.4.2 to 8.5.0 (#2813)
Bumps [tenacity](https://github.com/jd/tenacity) from 8.4.2 to 8.5.0.
- [Release notes](https://github.com/jd/tenacity/releases)
- [Commits](https://github.com/jd/tenacity/compare/8.4.2...8.5.0)

---
updated-dependencies:
- dependency-name: tenacity
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: yufansong <yufan@risingwave-labs.com>
2024-07-05 16:26:08 +00:00
மனோஜ்குமார் பழனிச்சாமி ed45a9e7b1 delete colima profile (#2807) 2024-07-05 08:55:26 -07:00
dependabot[bot] 1bdfbedccc chore(deps-dev): bump ruff from 0.5.0 to 0.5.1 (#2811)
Bumps [ruff](https://github.com/astral-sh/ruff) from 0.5.0 to 0.5.1.
- [Release notes](https://github.com/astral-sh/ruff/releases)
- [Changelog](https://github.com/astral-sh/ruff/blob/main/CHANGELOG.md)
- [Commits](https://github.com/astral-sh/ruff/compare/0.5.0...0.5.1)

---
updated-dependencies:
- dependency-name: ruff
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-05 08:46:42 -07:00
dependabot[bot] 274464101f chore(deps): bump litellm from 1.41.6 to 1.41.7 (#2812)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.41.6 to 1.41.7.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.41.6...v1.41.7)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-05 08:46:34 -07:00
Graham Neubig a081935fd8 Simplify eval code (#2775)
* Start simplifying eval code

* Update

* Add EDA

* Updated GAIA

* Update gpqa

* Add humanevalfix

* Fix logic_reasoning

* Add miniwob

* Add mint and ml_bench

* toolqa

* Added swe-bench

* Fixed webarena

* Refactor parameters
2024-07-05 19:33:08 +09:00
r.e.e.c.h.e.e 038e8f8caa docs: update docker run command to pull default 'latest' tag (#2804)
- Ensure users get the most recent stable release version when pulling default image.
- Explains the main tag for those who want the most recent updates.
2024-07-05 04:13:35 +00:00
மனோஜ்குமார் பழனிச்சாமி 143f38d25a Refactored sandbox config and added fast boot (#2455)
* Refactored sandbox config and added fastboot

* added tests

* fixed tests

* fixed tests

* intimate user about breaking change

* remove default config from eval

* check for lowercase env

* add test

* Revert Migration

* migrate old sandbox configs

* resolve merge conflict

* revert migration 2

* Revert "remove default config from eval"

This reverts commit de57c588db.

* change type to box_type

* fix var name

* linted

* lint

* lint comments

* fix tests

* fix tests

* fix typo

* fix box_type, remove fast_boot

* add tests for sandbox config

* fix test

* update eval docs

* small removal comments

* adapt toml template

* old fields shouldn't be in the app dataclass

* fix old keys in app config

* clean up exec box

---------

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
2024-07-05 03:30:21 +00:00
Xingyao Wang 82f4860470 retry for internal server error (#2806) 2024-07-05 01:47:55 +00:00
Xingyao Wang 298956c78a [Eval] initialize llm inside process_instance to circumvent "AttributeError:… (#2805)
* initialize llm inside process_instance to circumvent "AttributeError: Can't pickle local object"

* update kwargs
2024-07-05 01:26:03 +00:00
Xingyao Wang 0d3b3ffbf8 [Arch] Removing docker exec box (#2802)
* depracting docker exec box

* remove doc exec from workflow and docs
2024-07-04 23:15:25 +00:00
Xingyao Wang e6cdf18d3b [Evaluation] Log empty patch stats for SWE-Bench (#2776)
* bump swebench version since the fix PR is merged

* add empy generation stats from latest pr

* delete eval_outputs if it already exists

* handle non string patch
2024-07-05 07:03:27 +08:00
Engel Nyst 0b8d357bef Add event synchronously (#2700)
* add to event stream sync

* remove async from tests
2024-07-05 00:15:51 +02:00
sven 1b10e2b9d5 Make CodeAct finish task (#2673)
* Added feature to CodeAct agent to finish action instead of waiting for user input.

* Minor change

* Update agenthub/codeact_agent/codeact_agent.py

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>

* updated integration tests with claude-sonnet-3.5

* Update agenthub/codeact_agent/prompt.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* updated tests to remove typo in prompt

* resolve merge conflicts II

* revert unintended change of regenerate script

* re-regenerating prompts to resolve merge conflicts

---------

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-07-04 11:21:46 -07:00
மனோஜ்குமார் பழனிச்சாமி 688bd2a8fc Added local ollama models (#2433)
* added local ollama models

* add ollama_base_url config

* Update listen.py

* add docs

* Update opendevin/server/listen.py

Co-authored-by: Graham Neubig <neubig@gmail.com>

* lint

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-07-04 15:56:26 +00:00
dependabot[bot] 6853cbb4f6 chore(deps): bump framer-motion from 11.2.12 to 11.2.13 in /frontend (#2793)
Bumps [framer-motion](https://github.com/framer/motion) from 11.2.12 to 11.2.13.
- [Changelog](https://github.com/framer/motion/blob/main/CHANGELOG.md)
- [Commits](https://github.com/framer/motion/compare/v11.2.12...v11.2.13)

---
updated-dependencies:
- dependency-name: framer-motion
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-04 23:41:47 +08:00
dependabot[bot] 14b4213acf chore(deps-dev): bump openai from 1.35.9 to 1.35.10 (#2789) 2024-07-04 12:37:08 +00:00
dependabot[bot] d145dd78a3 chore(deps-dev): bump typescript from 5.2.2 to 5.5.3 in /docs (#2785) 2024-07-04 20:02:11 +08:00
dependabot[bot] c8270013ad chore(deps): bump react-dom from 18.2.0 to 18.3.1 in /docs (#2787) 2024-07-04 11:36:57 +00:00
dependabot[bot] dd1bd9caf3 chore(deps): bump react from 18.2.0 to 18.3.1 in /docs (#2786) 2024-07-04 19:03:46 +08:00
dependabot[bot] c77480bb55 chore(deps): bump litellm from 1.41.3 to 1.41.6 (#2790)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.41.3 to 1.41.6.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.41.3...v1.41.6)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-04 11:01:48 +00:00
dependabot[bot] 7291c320fc chore(deps): bump boto3 from 1.34.138 to 1.34.139 (#2788)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.138 to 1.34.139.
- [Release notes](https://github.com/boto/boto3/releases)
- [Commits](https://github.com/boto/boto3/compare/1.34.138...1.34.139)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-04 10:58:40 +00:00
Leo c2f557edde refactor: multiple code improvements (#2771) 2024-07-04 18:51:22 +08:00
Leo 869997b941 chore: Add docs for dependabot and add the open pr limit from 5 to 20. (#2784)
Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-04 18:07:30 +08:00
r.e.e.c.h.e.e d894347f9b Add 'latest' tag to Docker builds for stable releases only (#2781)
- Ensure 'latest' always points to the most recent stable version
- Address issue #2730: Release "latest" tag when pushing image to DockerHub
2024-07-04 18:06:54 +08:00
Leo 90a68ca816 chore: Add architecture diagram. (#2783)
* chore: Add architecture diagram.

Signed-off-by: ifuryst <ifuryst@gmail.com>

* Fix syntax error.

Signed-off-by: ifuryst <ifuryst@gmail.com>

---------

Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-04 18:06:14 +08:00
dependabot[bot] 82ed0a07cb chore(deps): bump jose from 5.6.2 to 5.6.3 in /frontend (#2766) 2024-07-04 13:28:35 +08:00
dependabot[bot] 81226e63e9 chore(deps): bump vite from 5.3.2 to 5.3.3 in /frontend (#2767) 2024-07-04 13:28:25 +08:00
dependabot[bot] 8b01d16926 chore(deps): bump react-router-dom from 6.24.0 to 6.24.1 in /frontend (#2768) 2024-07-04 13:28:13 +08:00
John Yang 89a3752c8c Restore SWE-bench dep refs (#2752)
* Restore SWE-bench dep refs

* update poetry lock

---------

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
2024-07-03 18:44:01 +00:00
Leo b60a696296 Fix the FE failed unit test. (#2773)
Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-03 15:52:51 +00:00
dependabot[bot] 018f45893b chore(deps-dev): bump openai from 1.35.8 to 1.35.9 (#2770)
Bumps [openai](https://github.com/openai/openai-python) from 1.35.8 to 1.35.9.
- [Release notes](https://github.com/openai/openai-python/releases)
- [Changelog](https://github.com/openai/openai-python/blob/main/CHANGELOG.md)
- [Commits](https://github.com/openai/openai-python/compare/v1.35.8...v1.35.9)

---
updated-dependencies:
- dependency-name: openai
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-03 23:22:26 +08:00
dependabot[bot] 5123e1d8d1 chore(deps): bump boto3 from 1.34.137 to 1.34.138 (#2769)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.137 to 1.34.138.
- [Release notes](https://github.com/boto/boto3/releases)
- [Commits](https://github.com/boto/boto3/compare/1.34.137...1.34.138)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-03 23:21:42 +08:00
r.e.e.c.h.e.e 94d68ca0b8 docs: Update custom sandbox guide to include steps to use pre-existin… (#2740)
* docs: Update custom sandbox guide to include steps to use pre-existing Docker images #2734

* docs: Update custom and pre-existing sandbox guide

* docs: Update custom and pre-existing sandbox guide
2024-07-03 12:08:55 +00:00
Leo 4dc01a7369 feature: Enable DEBUG level logging based on config setting. (#2762)
Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-03 20:21:41 +09:00
Graham Neubig ffd3c7144c Remove global args (#2760)
* Remove global args

* Remove global args

* Update files

* Update main

* Bug fixes

* Fix logging
2024-07-03 20:07:52 +09:00
Leo 0d0e6db1e3 feature: Add config template. (#2736)
Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-03 13:31:09 +09:00
Xingyao Wang 4d0c4f37d6 [Evaluation] fix SWE-Bench docker image name (#2751)
* fix double underscore

* remove unused script
2024-07-03 04:30:38 +08:00
dependabot[bot] f293b33856 chore(deps-dev): bump @typescript-eslint/eslint-plugin in /frontend (#2742)
Bumps [@typescript-eslint/eslint-plugin](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/eslint-plugin) from 7.14.1 to 7.15.0.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/eslint-plugin/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.15.0/packages/eslint-plugin)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/eslint-plugin"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-02 16:08:09 +00:00
dependabot[bot] 50efea5960 chore(deps): bump boto3 from 1.34.136 to 1.34.137 (#2744)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.136 to 1.34.137.
- [Release notes](https://github.com/boto/boto3/releases)
- [Commits](https://github.com/boto/boto3/compare/1.34.136...1.34.137)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-02 23:35:35 +08:00
dependabot[bot] 52ab4b4f18 chore(deps-dev): bump @typescript-eslint/parser in /frontend (#2743)
Bumps [@typescript-eslint/parser](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/parser) from 7.14.1 to 7.15.0.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/parser/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.15.0/packages/parser)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/parser"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-02 23:35:13 +08:00
dependabot[bot] 8ff9332358 chore(deps-dev): bump openai from 1.35.7 to 1.35.8 (#2745)
Bumps [openai](https://github.com/openai/openai-python) from 1.35.7 to 1.35.8.
- [Release notes](https://github.com/openai/openai-python/releases)
- [Changelog](https://github.com/openai/openai-python/blob/main/CHANGELOG.md)
- [Commits](https://github.com/openai/openai-python/compare/v1.35.7...v1.35.8)

---
updated-dependencies:
- dependency-name: openai
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-02 23:34:47 +08:00
dependabot[bot] c052ea30cc chore(deps): bump litellm from 1.40.29 to 1.41.3 (#2746)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.29 to 1.41.3.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.29...v1.41.3)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-02 23:34:11 +08:00
dependabot[bot] 4489c40106 chore(deps-dev): bump typescript from 5.5.2 to 5.5.3 in /frontend (#2741)
Bumps [typescript](https://github.com/Microsoft/TypeScript) from 5.5.2 to 5.5.3.
- [Release notes](https://github.com/Microsoft/TypeScript/releases)
- [Changelog](https://github.com/microsoft/TypeScript/blob/main/azure-pipelines.release.yml)
- [Commits](https://github.com/Microsoft/TypeScript/compare/v5.5.2...v5.5.3)

---
updated-dependencies:
- dependency-name: typescript
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-02 23:33:34 +08:00
மனோஜ்குமார் பழனிச்சாமி cc8204e5c1 CI: For colima, modify CPU count and memory (#2712)
* CI: For colima, modify CPUcount and memory

* fix arg
2024-07-02 15:44:44 +02:00
r.e.e.c.h.e.e 0689eef844 Fix: Add scroll functionality to file explorer sidepane (#2731)
* Fix: Add scroll and resize functionalities to file explorer

* fix: the icon missing from filename and tree scroll
2024-07-02 12:14:12 +00:00
PierrunoYT bfa1de4a6b Refactor: Enhance file handling and code editing functionality (#2646)
* refactor: Enhance file handling and code editing functionality

# PR Summary
**refactor: Enhance file handling and code editing functionality**

## PR Description

This pull request includes improvements to file handling, error management, and code editing functionality across multiple files. The changes enhance the robustness, security, and user experience of the application.

### Changes in `listen.py`

1. **Imports and Error Handling**:
   - Removed `warnings` import and its usage with `litellm`.
   - More consistent use of `JSONResponse` and `HTTPException` for error handling.

2. **WebSocket Endpoint (`/ws`)**:
   - Simplified logic for handling events using a single `isinstance` check.

3. **New Endpoint**:
   - Added `/api/save-file` POST endpoint for saving file contents.
   - Implemented checks for agent state before allowing file edits.

4. **Code Style and Organization**:
   - Improved code formatting and organization.
   - Refactored some functions for better readability and consistency.

### Changes in `fileService.ts`

1. **Error Handling**:
   - Added try-catch blocks to all functions for better error handling and logging.

2. **Input Sanitization**:
   - Implemented `encodeURIComponent()` for file names and paths in API requests.

3. **Type Checking**:
   - Added type checks for API responses to ensure data format consistency.

4. **File Upload Improvement**:
   - Refactored `uploadFiles()` to use `Array.from(files)` instead of a for loop.

5. **New Functionality**:
   - Added `saveFile()` function to allow saving file content to a specified path.

### Changes in `CodeEditor.tsx`

1. **New Dependencies**:
   - Added imports for state management, UI components, and file operations.

2. **State Management**:
   - Introduced new state variables for tracking save status and last saved time.
   - Implemented Redux state management for code and agent state.

3. **UI Enhancements**:
   - Added a save button with dynamic colors based on save status.
   - Implemented a save notification system.
   - Added a "Last saved" timestamp display.

4. **File Saving Functionality**:
   - Implemented complete file saving feature with error handling and user feedback.

5. **Code Structure**:
   - Improved structure with additional hooks and memoized values for optimization.

### Testing Performed
- Manually tested new file saving functionality.
- Verified error handling and user feedback mechanisms.
- Checked integration between backend (`listen.py`) and frontend (`fileService.ts`, `CodeEditor.tsx`).

### Next Steps
- Conduct thorough testing of the file saving feature across different scenarios.
- Update documentation to reflect new file handling capabilities.
- Consider adding unit tests for new functions and components.

* Added Docstrings back

Added Docstrings back

* Fix

# Allow Code Editing in AWAITING_USER_INPUT State

## Description
This pull request extends the functionality of the code editor to allow editing when the agent is in the AWAITING_USER_INPUT state, in addition to the existing PAUSED and FINISHED states.

## Changes
1. Backend (`listen.py`):
   - Updated the `save_file` function to allow saving when the agent state is AWAITING_USER_INPUT.

2. Frontend (`CodeEditor.tsx`):
   - Modified the `isEditingAllowed` condition to include the AWAITING_USER_INPUT state.

## Files Changed
- `listen.py`
- `CodeEditor.tsx`

## Testing
- Verified that the save button appears when the agent is in the AWAITING_USER_INPUT state.
- Tested saving files in all three allowed states (PAUSED, FINISHED, AWAITING_USER_INPUT).
- Ensured that saving is still prohibited in other agent states.

## Additional Notes
This change improves the user experience by allowing code edits while the agent is waiting for user input, which is a common scenario in interactive coding sessions.

* Add internationalization for 'File saved successfully' message

# Add internationalization for 'File saved successfully' message

## Description
This PR adds internationalization support for the "File saved successfully" message in the CodeEditor component. It updates the translation.json file to include translations for multiple languages and modifies the CodeEditor.tsx file to use the new translation key.

## Changes
1. Updated `translation.json`:
   - Added a new key `CODE_EDITOR$FILE_SAVED_SUCCESSFULLY` with translations for multiple languages.
   - Ensured the file structure supports multiple languages per key.

2. Modified `CodeEditor.tsx`:
   - Updated the success message to use the new translation key.
   - Applied the translation to both the toast notification and the on-screen notification.

## Why
These changes improve the user experience for non-English speakers by providing localized feedback when a file is successfully saved. This aligns with our goal of making the application more accessible to a global audience.

## How to Test
1. Change the application language to different supported languages.
2. Open the CodeEditor, make changes to a file, and save it.
3. Verify that the "File saved successfully" message appears in the correct language for both the toast and on-screen notifications.

## Additional Notes
Please pay special attention to the structure of the translation.json file to ensure it follows our established patterns for internationalization.

* Add toast notifications for error handling in fileService

# Add toast notifications for error handling in fileService

## Description
This PR enhances the error handling in the `fileService.ts` file by adding toast notifications for user feedback. It maintains the existing console error logging for debugging purposes while improving the user experience by providing visible error messages in the UI.

## Changes
- Added import for the toast utility
- Implemented toast.error() calls in catch blocks for all file operations
- Kept console.error() calls for detailed logging
- Updated error messages to be more user-friendly

## Files Changed
- `src/services/fileService.ts`

## Testing
- Tested all file operations (select, upload, list, save) to ensure proper error handling
- Verified that toast notifications appear when errors are simulated
- Confirmed that console errors are still logged for debugging

## Additional Notes
This change improves error visibility for users without altering the underlying error handling logic. It should make troubleshooting easier for both users and developers.

* Add file path safety check and improve error handling in file services

# Add file path safety check and improve error handling in file services

## Description
This PR enhances the `fileService.ts` by adding a safety check for file paths in the `saveFile` function and improves error handling across all file operations. It also includes new translations for various file-related error messages.

## Changes
1. Updated `src/services/fileService.ts`:
   - Added a validation check for file paths in the saveFile function
   - Improved error handling for all file operations (select, upload, list, save)
   - Implemented toast error messages with translation support

2. Updated `src/i18n/translations.json`:
   - Added new translation keys for file service error messages:
     - FILE_SERVICE$SELECT_FILE_ERROR
     - FILE_SERVICE$UPLOAD_FILES_ERROR
     - FILE_SERVICE$LIST_FILES_ERROR
     - FILE_SERVICE$SAVE_FILE_ERROR
     - FILE_SERVICE$INVALID_FILE_PATH

## Files Changed
- `src/services/fileService.ts`
- `src/i18n/translations.json`

## Key Implementation Details
```typescript
export async function saveFile(filePath: string, content: string): Promise<void> {
  const { t } = useTranslation();
  if (!filePath || filePath.includes('..')) {
    toast.error(t(I18nKey.FILE_SERVICE$INVALID_FILE_PATH));
    throw new Error('Invalid file path');
  }
  try {
    // Existing implementation...
  } catch (error) {
    console.error('Error saving file:', error);
    toast.error(t(I18nKey.FILE_SERVICE$SAVE_FILE_ERROR), 'File Save Error');
    throw error;
  }
}
```

## Testing
- Verified that the saveFile function rejects invalid file paths (empty or containing '..')
- Confirmed that appropriate error messages are displayed using toast notifications for all file operations
- Tested with different languages to ensure translated messages appear correctly

## Security Implications
The file path check in saveFile enhances security by preventing potential directory traversal attacks.

## Next Steps
- Consider adding similar safety checks to other file operations if applicable
- Ensure thorough testing of error scenarios across all supported languages

* Add docstrings to listen.py

# Add docstrings to listen.py

## Description
This PR adds comprehensive docstrings to all functions in the `listen.py` file. These additions improve code documentation, making the file more readable and maintainable for current and future developers.

## Changes
- Added docstrings to all functions in `listen.py`
- Docstrings follow the Google Python Style Guide format
- Included descriptions, parameters, return values, and potential exceptions for each function

## Files Changed
- `src/listen.py`

## Docstring Example
Here's an example of one of the added docstrings:

```python
@app.post('/api/save-file')
async def save_file(request: Request):
    """
    Save a file to the agent's runtime file store.

    This endpoint allows saving a file when the agent is in a paused, finished,
    or awaiting user input state. It checks the agent's state before proceeding
    with the file save operation.

    Args:
        request (Request): The incoming FastAPI request object.

    Returns:
        JSONResponse: A JSON response indicating the success of the operation.

    Raises:
        HTTPException:
            - 403 error if the agent is not in an allowed state for editing.
            - 400 error if the file path or content is missing.
            - 500 error if there's an unexpected error during the save operation.
    """
    # Function implementation...
```

## Impact
- Improved code readability and maintainability
- Better understanding of function purposes, inputs, outputs, and potential errors
- Easier onboarding for new developers working on this file
- Enhanced IDE support for function descriptions and parameter information

## Testing
- No functional changes were made, so existing tests should pass without modification
- Manual review of docstrings for accuracy and completeness is recommended

## Next Steps
- Consider adding similar docstrings to other files in the project for consistency
- Review the added docstrings to ensure they accurately describe the current functionality
- Update docstrings as needed when function implementations change in the future

## Additional Notes
The existing code structure and functionality remain unchanged. This PR focuses solely on improving documentation through the addition of docstrings.

* Revert exclude_list formatting and add docstrings in listen.py

# Revert exclude_list formatting and add docstrings in listen.py

## Description
This PR makes two main changes to the `listen.py` file:
1. Reverts the `exclude_list` in the `list_files` function to its original format, with each item on a separate line.
2. Adds comprehensive docstrings to all functions in the file.

These changes improve code readability, maintain consistency with project standards, and enhance documentation for better maintainability.

## Changes
1. Updated `opendevin/server/listen.py`:
   - Reverted `exclude_list` formatting in `list_files` function
   - Added docstrings to all functions

## Detailed Changes

### 1. Reverted exclude_list formatting
```python
exclude_list = (
    '.git',
    '.DS_Store',
    '.svn',
    '.hg',
    '.idea',
    '.vscode',
    '.settings',
    '.pytest_cache',
    '__pycache__',
    'node_modules',
    'vendor',
    'build',
    'dist',
    'bin',
    'logs',
    'log',
    'tmp',
    'temp',
    'coverage',
    'venv',
    'env',
)
```

### 2. Added docstrings (example)
```python
@app.get('/api/list-files')
def list_files(request: Request, path: str = '/'):
    """
    List files in the specified path.

    This function retrieves a list of files from the agent's runtime file store,
    excluding certain system and hidden files/directories.

    Args:
        request (Request): The incoming request object.
        path (str, optional): The path to list files from. Defaults to '/'.

    Returns:
        list: A list of file names in the specified path.

    Raises:
        HTTPException: If there's an error listing the files.
    """
    # Function implementation...
```

## Rationale
- Reverting `exclude_list` formatting maintains consistency with the project's coding style and ensures proper functioning of pre-commit hooks.
- Adding docstrings improves code documentation, making it easier for developers to understand and maintain the codebase.

## Impact
- Improved code readability and consistency
- Enhanced documentation for all functions in `listen.py`
- Easier onboarding for new developers
- Better IDE support for function descriptions and parameter information

## Testing
- No functional changes were made, so existing tests should pass without modification
- Manual review of the reverted `exclude_list` and new docstrings is recommended

## Additional Notes
- The existing code functionality remains unchanged
- All functions in `listen.py` now have detailed docstrings following the Google Python Style Guide format

## Next Steps
- Review the added docstrings to ensure they accurately describe the current functionality
- Consider adding similar docstrings to other files in the project for consistency
- Update docstrings as needed when function implementations change in the future

* made code reviewable

* fixed ruff issues

* Update listen.py docstrings

* final tweaks

* re-added encodedURIComponent in selectFile

---------

Co-authored-by: tobitege <tobitege@gmx.de>
Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com>
2024-07-02 09:42:43 +02:00
Leo 5e6fb6131f refactor: Renamed variables to resolve naming conflicts and eliminate warnings (#2732)
* refactor: Renamed variables to resolve naming conflicts and eliminate warnings

Signed-off-by: ifuryst <ifuryst@gmail.com>

* Fix lint failed.

Signed-off-by: ifuryst <ifuryst@gmail.com>

* Combine set_initial_state methods, rename _filed to f, and adjust the AppConfig update codes.

Signed-off-by: ifuryst <ifuryst@gmail.com>

---------

Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-02 15:00:58 +08:00
Xingyao Wang 41ddba84bd [Agent] (Potentially) improve Editing using diff (#2685)
* add replace-based block edit & preliminary test case fix

* further fix the insert behavior

* make edit only work on first occurence

* bump codeact version since we now use new edit agentskills

* update prompt for new agentskills

* update integration tests

* make run_infer.sh executable

* remove code block for edit_file

* update integration test for prompt changes

* default to not use hint for eval

* fix insert emptyfile bug

* throw value error when `to_replace` is empty

* make `_edit_or_insert_file` return string so we can try to fix some linter errors (best attempt)

* add todo

* update integration test

* fix sandbox test for this PR
2024-07-02 11:50:15 +09:00
Xingyao Wang 6a0ffc5c61 [Evaluation] Use the latest official SWE-Bench Dockerization for evaluation (#2728)
* add newline after patch to fix patch apply

* new swebench wip

* add newline after patch to fix patch apply

* only add newline if not empty

* update swebench source and update

* update gitignore for swebench eval

* update old prep_eval

* update gitignore

* add scripts for push and pull swebench images

* update eval_infer.sh

* update eval_infer for new docker workflow

* update script to create markdown report based on report.json

* update eval infer to use update output

* update readme

* only move result to folder if running whole file

* remove set-x

* update conversion script

* Update evaluation/swe_bench/README.md

* Update evaluation/swe_bench/README.md

* Update evaluation/swe_bench/README.md

* make sure last line end with newline

* switch to an fix attempt branch of swebench

* Update evaluation/swe_bench/README.md

* Update evaluation/swe_bench/README.md

---------

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
2024-07-01 23:58:30 +00:00
tobitege 6246cb8e67 fix: improve exit_code processing (ssh_box) (#2726)
* improve exit_code processing

* removed debug spam
2024-07-01 19:59:31 +00:00
Arno.Edwards c224a6f3f6 fix(docs): translate missing parts (#2724)
- fix incorrect format in sidebar.json
- add custom_sandbox_guide.md to other language folders
2024-07-01 18:57:21 +02:00
Leo 82ea8b3556 test: fix the failed unit tests. (#2721)
Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-01 15:27:27 +00:00
dependabot[bot] c88b9f8f6d chore(deps-dev): bump ruff from 0.4.10 to 0.5.0 (#2719)
Bumps [ruff](https://github.com/astral-sh/ruff) from 0.4.10 to 0.5.0.
- [Release notes](https://github.com/astral-sh/ruff/releases)
- [Changelog](https://github.com/astral-sh/ruff/blob/main/CHANGELOG.md)
- [Commits](https://github.com/astral-sh/ruff/compare/v0.4.10...0.5.0)

---
updated-dependencies:
- dependency-name: ruff
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-01 23:22:13 +08:00
dependabot[bot] e83d1fe8b8 chore(deps): bump @reduxjs/toolkit from 2.2.5 to 2.2.6 in /frontend (#2717)
Bumps [@reduxjs/toolkit](https://github.com/reduxjs/redux-toolkit) from 2.2.5 to 2.2.6.
- [Release notes](https://github.com/reduxjs/redux-toolkit/releases)
- [Commits](https://github.com/reduxjs/redux-toolkit/compare/v2.2.5...v2.2.6)

---
updated-dependencies:
- dependency-name: "@reduxjs/toolkit"
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-01 23:21:47 +08:00
dependabot[bot] 1cbbb10c56 chore(deps-dev): bump postcss from 8.4.38 to 8.4.39 in /frontend (#2716)
Bumps [postcss](https://github.com/postcss/postcss) from 8.4.38 to 8.4.39.
- [Release notes](https://github.com/postcss/postcss/releases)
- [Changelog](https://github.com/postcss/postcss/blob/main/CHANGELOG.md)
- [Commits](https://github.com/postcss/postcss/compare/8.4.38...8.4.39)

---
updated-dependencies:
- dependency-name: postcss
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-01 23:21:24 +08:00
dependabot[bot] 51a71aaec9 chore(deps): bump boto3 from 1.34.135 to 1.34.136 (#2718)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.135 to 1.34.136.
- [Release notes](https://github.com/boto/boto3/releases)
- [Commits](https://github.com/boto/boto3/compare/1.34.135...1.34.136)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-07-01 15:18:28 +00:00
Leo 58191e0af0 style: refine the copy button and add copy feedback for the icon. (#2715)
* Add padding on top for input to overlay the message bubbles.

Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-07-01 14:59:12 +00:00
tobitege 7fda69ff10 mixin: improve logging (#2713)
* mixin: improve logging

* refactor logger creation
2024-07-01 14:58:42 +02:00
tobitege ea3c58dc8c feat: added make run-wsl (#2711)
* feat: added make run-wsl

* moved Windows warning
2024-07-01 09:01:54 +02:00
Engel Nyst 1975689cd4 remove swe agent (#2708) 2024-07-01 12:27:14 +09:00
mamoodi 00542432bd docs: Update documentation with some consistency (#2706)
* Update documentation with some consistency

* Make windows troubleshooting a little more clear

* Apply suggestions from code review

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

---------

Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-30 23:27:58 +00:00
Boxuan Li 8dae1f9307 Bypass MAX_ITERATIONS and MAX_BUDGET_PER_TASK on web GUI (#2697)
Closes #1493

Introduced TRAFFIC_CONTROL_STATE to allow OpenDevin to switch between normal traffic limiting mode and temporarily disabled mode.
2024-06-30 13:19:45 -07:00
Engel Nyst e24c52d060 Small refactoring of obs truncation (#2701)
* refactor truncate_content a bit to be usable by all agents

* adjust doc
2024-06-30 12:12:08 +02:00
Engel Nyst 2d9bb56763 Add ability to restore the cli session (optional) (#2699)
* add ability to restore the main session

* add quick log

* rename to cli session
2024-06-30 06:56:55 +00:00
Engel Nyst 874b4c9075 CLI concurrency (#2695)
* add session id in cli, evals

* fix main sid
2024-06-30 04:04:30 +02:00
Xingyao Wang 15e0c524f4 default to not use hint for eval (#2696) 2024-06-29 21:27:57 +00:00
Engel Nyst 4b1cc56682 Sync history to stream (#2640)
* add event to stream before budget check

* make the budget check before the step

* Update opendevin/controller/agent_controller.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-29 21:19:00 +00:00
மனோஜ்குமார் பழனிச்சாமி b4f63ae076 doc: Update CONTRIBUTING.md (#2671) 2024-06-29 22:06:14 +02:00
Boxuan Li e45b311c35 Remove MAX_CHARS traffic control (#2694)
* Remove MAX_CHARS limiting

* More cleanup
2024-06-29 12:59:41 -07:00
mamoodi 75f3181c08 Update tag to use in README and docs to 0.7.0 (#2683)
Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
2024-06-29 19:53:49 +00:00
Boxuan Li 79a3e81708 ghcr: Fix local built image name in tests (#2686)
* ghcr: Fix local built image name in tests

* Push to registry when tagged

* Fix push and experiment

* Remove debug
2024-06-29 09:53:30 -07:00
Xingyao Wang e8cb6803df [Evaluation] Improve patch apply in SWE-Bench (#2684)
* add newline after patch to fix patch apply

* only add newline if not empty
2024-06-29 14:11:07 +08:00
tobitege 7d31057904 feat: file explorer: better sorting; .gitignore support; file upload config (#2621)
* feat: file explorer: better sorting; .gitignore support; file upload config

* resolved poetry

* move config settings (no extra file); updated uploading of files; fix exception on refresh of removed folder

* removed console cmds; fix in a toast

* attempt fix of upload toasts

* fix new options' assignments in listen.py
2024-06-28 16:36:25 +00:00
dependabot[bot] b88898235e chore(deps-dev): bump llama-index-vector-stores-chroma (#2680) 2024-06-28 23:49:49 +08:00
dependabot[bot] a9302b15df chore(deps): bump json-repair from 0.25.1 to 0.25.2 (#2679) 2024-06-28 23:49:19 +08:00
dependabot[bot] 52c7019cb3 chore(deps): bump litellm from 1.40.28 to 1.40.29 (#2677) 2024-06-28 23:48:57 +08:00
dependabot[bot] b05aaffd7e chore(deps-dev): bump openai from 1.35.6 to 1.35.7 (#2676) 2024-06-28 23:48:37 +08:00
dependabot[bot] a918198f43 chore(deps): bump vite from 5.3.1 to 5.3.2 in /frontend (#2681) 2024-06-28 23:48:02 +08:00
dependabot[bot] 839c51ed3c chore(deps): bump jose from 5.6.1 to 5.6.2 in /frontend (#2682) 2024-06-28 23:47:35 +08:00
dependabot[bot] 105662a40c chore(deps): bump boto3 from 1.34.134 to 1.34.135 (#2678)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.134 to 1.34.135.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.134...1.34.135)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-28 15:08:16 +00:00
Boxuan Li 7766a3283e CodeActAgent: Fix delegate history (#2672) 2024-06-28 16:37:23 +09:00
மனோஜ்குமார் பழனிச்சாமி af9385322b Refactor: Simplify message formatting (#2670)
Removed redundant `str()` conversion in f-string.
2024-06-28 07:34:26 +02:00
Jiayi Pan 917d96e06f Fix doc error in evals (#2654) 2024-06-27 16:13:47 +00:00
dependabot[bot] cc9cb1bac0 chore(deps): bump litellm from 1.40.27 to 1.40.28 (#2664)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.27 to 1.40.28.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.27...v1.40.28)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-27 15:44:54 +00:00
dependabot[bot] fdef348365 chore(deps): bump jose from 5.5.0 to 5.6.1 in /frontend (#2663) 2024-06-27 15:42:38 +00:00
dependabot[bot] 1827579f4b chore(deps): bump google-generativeai from 0.7.0 to 0.7.1 (#2662) 2024-06-27 23:40:56 +08:00
dependabot[bot] c2969d422a chore(deps-dev): bump openai from 1.35.4 to 1.35.6 (#2660) 2024-06-27 23:40:28 +08:00
dependabot[bot] badfb054f1 chore(deps): bump boto3 from 1.34.133 to 1.34.134 (#2661) 2024-06-27 23:39:58 +08:00
மனோஜ்குமார் பழனிச்சாமி 9919d8e448 Provide [Package already installed] info to LLM (#2642)
* Provide [Package already installed] info to LLM

* regenerate tests
2024-06-27 09:03:54 +00:00
Engel Nyst 58b06cced7 Revert "Show relevant error in UI (#2516)" (#2657)
This reverts commit d0bdae232f.
2024-06-27 08:55:41 +00:00
Boxuan Li 5835680292 Add test for auto_lint after file edit (#2655) 2024-06-27 17:13:45 +09:00
மனோஜ்குமார் பழனிச்சாமி 877d9ae794 Added Sound Notification 🎵 (#2203)
* Added Sound Notification

* prettify

* linted

* disabled by default

* Add dev config for frontend

* add volume icon

* fix typo

* lint

* lint

* Apply suggestions from code review

Co-authored-by: Graham Neubig <neubig@gmail.com>

* Move volume icon to bottom

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-06-27 01:27:48 +00:00
Xingyao Wang 6099f72ae3 Update docusaurus.config.ts for a changed Doc URL (#2653)
* Update docusaurus.config.ts for a changed Doc URL

* fix url

* fix broken link again

* fix url again
2024-06-26 21:56:42 +00:00
Xingyao Wang 418cda23c5 Update custom_sandbox_guide.md (#2650) 2024-06-26 21:58:19 +02:00
dependabot[bot] 3fb8206b72 chore(deps-dev): bump openai from 1.35.3 to 1.35.4 (#2649)
Bumps [openai](https://github.com/openai/openai-python) from 1.35.3 to 1.35.4.
- [Release notes](https://github.com/openai/openai-python/releases)
- [Changelog](https://github.com/openai/openai-python/blob/main/CHANGELOG.md)
- [Commits](https://github.com/openai/openai-python/compare/v1.35.3...v1.35.4)

---
updated-dependencies:
- dependency-name: openai
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 15:51:57 +00:00
dependabot[bot] 7394c91c0e chore(deps): bump litellm from 1.40.26 to 1.40.27 (#2648)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.26 to 1.40.27.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.26...v1.40.27)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 15:46:56 +00:00
dependabot[bot] acee3e9d04 chore(deps): bump boto3 from 1.34.132 to 1.34.133 (#2647)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.132 to 1.34.133.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.132...1.34.133)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 15:42:58 +00:00
dependabot[bot] 8fef94cac4 chore(deps): bump jose from 5.4.1 to 5.5.0 in /frontend (#2645)
Bumps [jose](https://github.com/panva/jose) from 5.4.1 to 5.5.0.
- [Release notes](https://github.com/panva/jose/releases)
- [Changelog](https://github.com/panva/jose/blob/main/CHANGELOG.md)
- [Commits](https://github.com/panva/jose/compare/v5.4.1...v5.5.0)

---
updated-dependencies:
- dependency-name: jose
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 23:40:14 +08:00
Sheun Aluko, MD-MS 0e8d1e5fb8 added custom sandbox guide markdown file (#2637) 2024-06-26 23:33:53 +08:00
dependabot[bot] 10b7a1e9c5 chore(deps-dev): bump @types/node from 20.14.8 to 20.14.9 in /frontend (#2644)
Bumps [@types/node](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/node) from 20.14.8 to 20.14.9.
- [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases)
- [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/node)

---
updated-dependencies:
- dependency-name: "@types/node"
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 23:29:55 +08:00
dependabot[bot] 5abe014218 chore(deps): bump react-router-dom from 6.23.1 to 6.24.0 in /frontend (#2628)
Bumps [react-router-dom](https://github.com/remix-run/react-router/tree/HEAD/packages/react-router-dom) from 6.23.1 to 6.24.0.
- [Release notes](https://github.com/remix-run/react-router/releases)
- [Changelog](https://github.com/remix-run/react-router/blob/main/packages/react-router-dom/CHANGELOG.md)
- [Commits](https://github.com/remix-run/react-router/commits/react-router-dom@6.24.0/packages/react-router-dom)

---
updated-dependencies:
- dependency-name: react-router-dom
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 23:19:52 +08:00
tobitege 163ee3abf6 dev: added make-i18n to "build" (package.json) (#2641)
* added i18n to build-frontend

* revert Makefile; update package.json for make-i18n instead
2024-06-26 11:53:40 +02:00
Boxuan Li ee86d8d25e Frontend support for delegation and rejection (#2608)
1. Add support for rejection action on frontend
2. Show users the reason for rejection
3. Get rid of weird empty box after delegation
4. On web GUI, show customer when a delegation starts and ends
2024-06-26 00:30:10 -07:00
Xavier Vergés cd91d45b44 Allow SANDBOX_CONTAINER_IMAGEs built from opendevin/sandbox:main (#2622) 2024-06-26 12:05:07 +08:00
tobitege 94d24b71d7 frontend: apply more i18n (#2639) 2024-06-26 05:54:35 +02:00
PierrunoYT fcbf1f8242 feat(frontend): Add Typing Indicator for Agent Response using Tailwind CSS (#2630)
* feat(frontend): Add Typing Indicator for Agent Response

* finetuning and linting

* evil lil linter

* TailwindCSS styles now

* improve bounce

* merge fix

* merge fix 2

* one more fix

* poetry.lock fix

---------

Co-authored-by: tobitege <tobitege@gmx.de>
2024-06-26 04:15:04 +02:00
dependabot[bot] 7ff746250e chore(deps): bump boto3 from 1.34.131 to 1.34.132 (#2632)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.131 to 1.34.132.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.131...1.34.132)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 02:37:28 +02:00
dependabot[bot] 2b1d2104d9 chore(deps-dev): bump reportlab from 4.2.0 to 4.2.2 (#2635)
Bumps [reportlab](https://www.reportlab.com/) from 4.2.0 to 4.2.2.

---
updated-dependencies:
- dependency-name: reportlab
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 00:45:12 +08:00
dependabot[bot] 406bb13caf chore(deps): bump tenacity from 8.4.1 to 8.4.2 (#2631)
Bumps [tenacity](https://github.com/jd/tenacity) from 8.4.1 to 8.4.2.
- [Release notes](https://github.com/jd/tenacity/releases)
- [Commits](https://github.com/jd/tenacity/compare/8.4.1...8.4.2)

---
updated-dependencies:
- dependency-name: tenacity
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 00:44:52 +08:00
dependabot[bot] c1009069ab chore(deps-dev): bump @typescript-eslint/eslint-plugin in /frontend (#2629)
Bumps [@typescript-eslint/eslint-plugin](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/eslint-plugin) from 7.13.1 to 7.14.1.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/eslint-plugin/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.14.1/packages/eslint-plugin)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/eslint-plugin"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-25 16:33:12 +00:00
dependabot[bot] 7d413c288e chore(deps): bump litellm from 1.40.25 to 1.40.26 (#2633)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.25 to 1.40.26.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.25...v1.40.26)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 00:06:06 +08:00
dependabot[bot] 781a6250f8 chore(deps-dev): bump mypy from 1.10.0 to 1.10.1 (#2634)
Bumps [mypy](https://github.com/python/mypy) from 1.10.0 to 1.10.1.
- [Changelog](https://github.com/python/mypy/blob/master/CHANGELOG.md)
- [Commits](https://github.com/python/mypy/compare/v1.10.0...v1.10.1)

---
updated-dependencies:
- dependency-name: mypy
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-26 00:05:45 +08:00
dependabot[bot] ffda754256 chore(deps-dev): bump @typescript-eslint/parser in /frontend (#2627)
Bumps [@typescript-eslint/parser](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/parser) from 7.13.1 to 7.14.1.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/parser/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.14.1/packages/parser)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/parser"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-25 23:55:40 +08:00
dependabot[bot] 7ca7f803e1 chore(deps): bump framer-motion from 11.2.11 to 11.2.12 in /frontend (#2626)
Bumps [framer-motion](https://github.com/framer/motion) from 11.2.11 to 11.2.12.
- [Changelog](https://github.com/framer/motion/blob/main/CHANGELOG.md)
- [Commits](https://github.com/framer/motion/compare/v11.2.11...v11.2.12)

---
updated-dependencies:
- dependency-name: framer-motion
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-25 15:37:44 +00:00
tobitege dab9a60e00 feat: Agent buttons hover decor (#2623)
* feat: Agent buttons hover decor

* use TailwindCSS for styling
2024-06-25 15:44:34 +02:00
PierrunoYT c99b94dbe9 feat(frontend): Add "Copy" Button to Chat Messages (#2619)
* Add "Copy" Button to Chat Messages

### PR Overview: Add "Copy" Button to Chat Messages

**Description:**
This PR introduces a "Copy" button to each chat message in the `ChatMessage` component. The button allows users to easily copy the content of a chat message to their clipboard. The implementation includes a button with a clipboard icon and the necessary functionality to copy the message content.

**Changes Made:**
1. **Imports:**
   - Added `FaClipboard` from `react-icons/fa` for the clipboard icon.
   - Added `toast` from `#utils/toast` for displaying notifications.

2. **New Functionality:**
   - Implemented `copyToClipboard` function using `navigator.clipboard.writeText` to copy the message content.
   - Added a button element with an `onClick` handler to trigger the `copyToClipboard` function.

3. **UI Enhancements:**
   - The button is styled to match the existing UI and is placed next to the message content.

**Code Changes:**
- Modified `frontend/src/components/chat/ChatMessage.tsx` to include the new button and functionality.

**Testing:**
- Verified that clicking the "Copy" button copies the message content to the clipboard.
- Confirmed that a toast notification appears upon successful copy or failure.

**Example Code:**
```tsx
import React from "react";
import Markdown from "react-markdown";
import { twMerge } from "tailwind-merge";
import { code } from "../markdown/code";
import { FaClipboard } from "react-icons/fa";
import toast from "#/utils/toast"; // Assuming you have a toast utility for notifications

interface MessageProps {
  message: Message;
}

function ChatMessage({ message }: MessageProps) {
  const className = twMerge(
    "markdown-body",
    "p-3 text-white max-w-[90%] overflow-y-auto rounded-lg",
    message.sender === "user" ? "bg-neutral-700 self-end" : "bg-neutral-500",
  );

  const copyToClipboard = () => {
    navigator.clipboard.writeText(message.content).then(() => {
      toast.info("Message copied to clipboard!");
    }).catch((error) => {
      toast.error(`Failed to copy message: ${error}`);
    });
  };

  return (
    <div data-testid="message" className={className}>
      <div className="flex justify-between items-center">
        <Markdown components={{ code }}>{message.content}</Markdown>
        <button
          onClick={copyToClipboard}
          className="ml-2 p-1 bg-neutral-600 rounded hover:bg-neutral-500"
          aria-label="Copy message"
        >
          <FaClipboard />
        </button>
      </div>
    </div>
  );
}

export default ChatMessage;
```

**Notes:**
- Ensure that the `react-icons` package is installed (`npm install react-icons` or `yarn add react-icons`).
- The toast utility is assumed to be available for notifications. If not, consider using an alternative notification method.

* layout enhancements; linting

---------

Co-authored-by: tobitege <tobitege@gmx.de>
Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com>
2024-06-25 06:15:44 +00:00
Boxuan Li 7e78fde48f Bug fix: add error observation to history (#2610)
* Bug fix: add error observation to history

* Regenerate to demonstrate format error
2024-06-24 21:24:17 -07:00
dependabot[bot] e71b8d1464 chore(deps-dev): bump @types/node from 20.14.7 to 20.14.8 in /frontend (#2617)
Bumps [@types/node](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/node) from 20.14.7 to 20.14.8.
- [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases)
- [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/node)

---
updated-dependencies:
- dependency-name: "@types/node"
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-24 21:36:33 +03:00
dependabot[bot] 4b35497b54 chore(deps): bump litellm from 1.40.20 to 1.40.25 (#2615)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.20 to 1.40.25.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.20...v1.40.25)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-24 19:08:03 +02:00
dependabot[bot] 6042b226a0 chore(deps): bump boto3 from 1.34.130 to 1.34.131 (#2616)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.130 to 1.34.131.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.130...1.34.131)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-24 19:04:28 +02:00
tobitege 1117dfebeb feat: update version to 0.6.2. added Action to update pyproject on Release (#2552)
* updated version; added Action to update pyproject version by current tag (if changed)

* higer pyproject version creates a tag now

* Release-only run to write tag to pyproject
2024-06-24 18:34:57 +02:00
Boxuan Li 8bce806dce Tweak prompts of ManagerAgent and CommitWriterAgent (#2609)
* Tweak prompts of ManagerAgent and CommitWriterAgent

* Fix prompts
2024-06-24 00:14:28 -07:00
Boxuan Li 39d90c0b2a Track metrics throughout delegation & Polish UX for out of budget error (#2595)
* Track metrics (costs) throught delegation

* Metrics should be shared across agents for better UX

* Update cost before starting delegate
2024-06-23 18:38:52 -07:00
Xingyao Wang 6de584d77d update swe-bench output with eval results (#2606) 2024-06-24 08:07:28 +09:00
mamoodi 4d1ffa1aaf Default makefile for persist_sandbox to be false (#2605)
Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
2024-06-23 15:56:56 +00:00
Arno.Edwards e1caf1db33 Add i18n support for official website (#2463)
* feat(i18n): initial i18n setup

- Configured i18n settings in docusaurus.config.js
- Implemented Translate component and translate function in key components

* docs(i18n): complete documentation internationalization

- Added support for Simplified Chinese and French

* Update docs/i18n/zh-Hans/docusaurus-plugin-content-docs/current/usage/troubleshooting/troubleshooting.md

* Update docs/i18n/zh-Hans/docusaurus-plugin-content-docs/current/usage/troubleshooting/troubleshooting.md

* Update docs/i18n/zh-Hans/docusaurus-plugin-content-docs/current/usage/troubleshooting/troubleshooting.md

* fix(build): resolve broken links causing build failure

- Fix issue causing build errors due to broken links in Docusaurus documentation.
- Resolve uncontrolled resource consumption in braces (see: https://github.com/advisories/GHSA-grv7-fg5c-xmjg).
- Bump Docusaurus to ^3.4.0 to fix MDX loader: linkify should process the md AST instead of the md string.

* fix: sync with commit 868b746

-  Change to `docusaurus write-translations` to provide translation for JSON files.

---------

Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-06-23 15:13:04 +00:00
sp.wack 868b7460d4 Replace console errors with toast errors (#2601) 2024-06-23 07:09:16 +00:00
sp.wack 63a3285b23 Disable eslint rule that throws useeffect warning (#2600) 2024-06-23 06:32:37 +00:00
sp.wack 0c1d6f8b4e fix(frontend): Prevent actions from disappearing before sending data (#2599)
* Prevent actions from disappearing before sending data

* Lint
2024-06-23 06:28:48 +00:00
மனோஜ்குமார் பழனிச்சாமி c455a09e43 Remove Colima and lima directory after uninstalling for Mac OS CI (#2598)
* Remove colima dir after uninstall

* Delete lima dir
2024-06-23 08:28:29 +02:00
Graham Neubig cab7a288ca Add NUM_WORKERS variable to run_infer.sh scripts for configurable woker settings (#2597)
* Add NUM_WORKERS variable to run_infer.sh scripts for configurable worker settings

* Update evaluation/webarena/scripts/run_infer.sh

---------

Co-authored-by: OpenDevin <opendevin@all-hands.dev>
2024-06-23 03:43:43 +00:00
mamoodi 9af2a08eda Remove PERSIST_SANDBOX=true option (#2591) 2024-06-22 17:59:34 +00:00
மனோஜ்குமார் பழனிச்சாமி 6bf1b56f06 Revert "Enable "vz" vm-type for MacOS CI (#2586)" (#2588)
This reverts commit 57b56c0536.
2024-06-22 23:45:14 +08:00
Graham Neubig 6312c21f8e Update documentation regarding feedback data usage (#2585)
* Update feedback data usage

* Update hugging face
2024-06-22 23:39:40 +08:00
மனோஜ்குமார் பழனிச்சாமி 57b56c0536 Enable "vz" vm-type for MacOS CI (#2586) 2024-06-22 12:30:18 +00:00
மனோஜ்குமார் பழனிச்சாமி 4c5e5a0b5a Mention Ubuntu supported versions (#2584) 2024-06-22 10:19:52 +00:00
sp.wack fa86e32231 Update feedback modal content (#2582) 2024-06-22 08:12:39 +00:00
Graham Neubig 45d7a53b91 Add links to a feedback sharing site (#2580)
* Add links to a feedback sharing site

* Remove console log

---------

Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com>
2024-06-22 09:36:31 +02:00
dependabot[bot] 7ad6ab1548 chore(deps-dev): bump ruff from 0.4.9 to 0.4.10 (#2573)
Bumps [ruff](https://github.com/astral-sh/ruff) from 0.4.9 to 0.4.10.
- [Release notes](https://github.com/astral-sh/ruff/releases)
- [Changelog](https://github.com/astral-sh/ruff/blob/main/CHANGELOG.md)
- [Commits](https://github.com/astral-sh/ruff/compare/v0.4.9...v0.4.10)

---
updated-dependencies:
- dependency-name: ruff
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-06-22 03:14:30 +00:00
dependabot[bot] ec43e67ae0 chore(deps): bump litellm from 1.40.17 to 1.40.20 (#2576)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.17 to 1.40.20.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.17...v1.40.20)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-06-22 03:13:12 +00:00
மனோஜ்குமார் பழனிச்சாமி c743320201 Interactive Terminal (#2493)
* Interactive Terminal

* linted

* fixed tests

* fixed tests

* refactored logic

* remove console logs
2024-06-21 20:56:54 -06:00
மனோஜ்குமார் பழனிச்சாமி 0845d475b8 Fix Mac CI Test (#2569)
* Fix Mac CI Test

* Start colima service

* unlink colima dependency: go

* Check for colima

Co-authored-by: Graham Neubig <neubig@gmail.com>

* fix indent

Co-authored-by: Graham Neubig <neubig@gmail.com>

* Try with uninstall

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-06-22 02:36:23 +00:00
Shimada666 5972498e23 No longer chown -R the miniforge3 folder (#2566)
* No longer chown -R the miniforge3 folder

* change miniforge3 group permission
2024-06-21 17:33:00 +00:00
dependabot[bot] 4567dd60e4 chore(deps): bump json-repair from 0.25.0 to 0.25.1 (#2575)
Bumps [json-repair](https://github.com/mangiucugna/json_repair) from 0.25.0 to 0.25.1.
- [Release notes](https://github.com/mangiucugna/json_repair/releases)
- [Commits](https://github.com/mangiucugna/json_repair/compare/0.25.0...0.25.1)

---
updated-dependencies:
- dependency-name: json-repair
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-21 09:38:37 -06:00
dependabot[bot] 13adfe5625 chore(deps-dev): bump openai from 1.35.1 to 1.35.3 (#2574)
Bumps [openai](https://github.com/openai/openai-python) from 1.35.1 to 1.35.3.
- [Release notes](https://github.com/openai/openai-python/releases)
- [Changelog](https://github.com/openai/openai-python/blob/main/CHANGELOG.md)
- [Commits](https://github.com/openai/openai-python/compare/v1.35.1...v1.35.3)

---
updated-dependencies:
- dependency-name: openai
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-21 09:38:21 -06:00
dependabot[bot] 900743ee7a chore(deps-dev): bump streamlit from 1.35.0 to 1.36.0 (#2572)
Bumps [streamlit](https://github.com/streamlit/streamlit) from 1.35.0 to 1.36.0.
- [Release notes](https://github.com/streamlit/streamlit/releases)
- [Commits](https://github.com/streamlit/streamlit/compare/1.35.0...1.36.0)

---
updated-dependencies:
- dependency-name: streamlit
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-21 09:38:02 -06:00
dependabot[bot] 1076789c23 chore(deps-dev): bump @types/node from 20.14.6 to 20.14.7 in /frontend (#2571)
Bumps [@types/node](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/node) from 20.14.6 to 20.14.7.
- [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases)
- [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/node)

---
updated-dependencies:
- dependency-name: "@types/node"
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-21 15:23:25 +00:00
dependabot[bot] 3f158d09e2 chore(deps-dev): bump typescript from 5.4.5 to 5.5.2 in /frontend (#2570)
Bumps [typescript](https://github.com/Microsoft/TypeScript) from 5.4.5 to 5.5.2.
- [Release notes](https://github.com/Microsoft/TypeScript/releases)
- [Changelog](https://github.com/microsoft/TypeScript/blob/main/azure-pipelines.release.yml)
- [Commits](https://github.com/Microsoft/TypeScript/compare/v5.4.5...v5.5.2)

---
updated-dependencies:
- dependency-name: typescript
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-21 15:18:21 +00:00
Graham Neubig 7e44c8f348 Update doc to clarify OpenDevin mission and link directly to content (#2568)
* Update doc

* Fix review comments
2024-06-21 15:12:40 +00:00
tobitege 6879acdfe7 fix: Makefile shall pull sandbox:main, not :latest (#2561) 2024-06-21 11:27:56 +02:00
Shimada666 64c2a783d4 Revert "Always pull sandbox image (#2538)" (#2560)
This reverts commit 6dd2491944.
2024-06-21 08:09:26 +00:00
Shimada666 1ffaed48c4 Stop always pulling the latest image. (#2558)
LGTM, thanks!
2024-06-21 05:42:49 +00:00
Boxuan Li 01fa52d062 Enforce linter in tests folder (#2557) 2024-06-20 21:50:34 -07:00
மனோஜ்குமார் பழனிச்சாமி f3ba2f02d5 Fix Mac OS CI - usernet unable to resolve IP for SSH forwarding (#2556) 2024-06-21 10:02:26 +05:30
dependabot[bot] 2df629f912 chore(deps-dev): bump eslint-plugin-jsx-a11y in /frontend (#2543)
Bumps [eslint-plugin-jsx-a11y](https://github.com/jsx-eslint/eslint-plugin-jsx-a11y) from 6.8.0 to 6.9.0.
- [Release notes](https://github.com/jsx-eslint/eslint-plugin-jsx-a11y/releases)
- [Changelog](https://github.com/jsx-eslint/eslint-plugin-jsx-a11y/blob/main/CHANGELOG.md)
- [Commits](https://github.com/jsx-eslint/eslint-plugin-jsx-a11y/compare/v6.8.0...v6.9.0)

---
updated-dependencies:
- dependency-name: eslint-plugin-jsx-a11y
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Leo <ifuryst@gmail.com>
2024-06-21 07:30:29 +03:00
மனோஜ்குமார் பழனிச்சாமி 41564c2eac Use :main instead of :latest (#2539)
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-21 03:57:50 +00:00
tobitege ff9f34bea2 sec: update npm module to 8.17.1 (#2554) 2024-06-21 02:45:09 +00:00
Boxuan Li bfa00932cc Enable test_agnostic_sandbox_jupyter_agentskills_fileop_pwd in CI (#2534)
* Enable test_agnostic_sandbox_jupyter_agentskills_fileop_pwd in CI

* Fix env variable value
2024-06-20 20:39:11 -06:00
Boxuan Li feabc97aba Evaluation time travel: build sandbox on the fly (#2491) 2024-06-20 20:22:02 -06:00
மனோஜ்குமார் பழனிச்சாமி 6dd2491944 Always pull sandbox image (#2538) 2024-06-21 06:52:18 +05:30
dependabot[bot] 0b742826ea chore(deps-dev): bump openai from 1.34.0 to 1.35.1 (#2547) 2024-06-21 01:17:51 +08:00
Xingyao Wang eac05d71fa replace / for tag (#2550) 2024-06-20 11:09:29 -06:00
Shimada666 0c2882d552 add workspace text (#2548) 2024-06-21 00:57:10 +08:00
மனோஜ்குமார் பழனிச்சாமி 35fc2177d0 Fix Mac CI test (#2544) 2024-06-21 00:56:23 +08:00
dependabot[bot] 8e22cd5f5e chore(deps-dev): bump @types/node from 20.14.5 to 20.14.6 in /frontend (#2542) 2024-06-21 00:36:46 +08:00
dependabot[bot] cac8159fd8 chore(deps): bump boto3 from 1.34.129 to 1.34.130 (#2546) 2024-06-21 00:36:16 +08:00
dependabot[bot] ff0a35451d chore(deps): bump monaco-editor from 0.49.0 to 0.50.0 in /frontend (#2541)
Bumps [monaco-editor](https://github.com/microsoft/monaco-editor) from 0.49.0 to 0.50.0.
- [Release notes](https://github.com/microsoft/monaco-editor/releases)
- [Changelog](https://github.com/microsoft/monaco-editor/blob/main/CHANGELOG.md)
- [Commits](https://github.com/microsoft/monaco-editor/compare/v0.49.0...v0.50.0)

---
updated-dependencies:
- dependency-name: monaco-editor
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-20 09:52:26 -06:00
Robert Brennan 373a700599 Architecture documentation (#2116)
* add initial arch

* add agent

* update docs

* Update opendevin/README.md

* Update opendevin/README.md

* Update opendevin/README.md
2024-06-20 15:16:06 +00:00
Shimada666 9ec5e4f3e2 remove gcc (#2536) 2024-06-20 13:49:26 +05:30
மனோஜ்குமார் பழனிச்சாமி 833bb50fb0 Downgrade Mac version in CI/CD Pipeline (#2499)
* downgrade mac version in CI

* Delete run-integration-tests.yml
2024-06-19 22:12:15 -07:00
Shimada666 26fc3c886a Make plugins sandbox-agnostic (#2101)
* tmp

* tmp

* merge main

* feat: auto build image cache

* remove plugins

* use config file

* update mamba setup shell

* support agnostic sandbox image autobuild

* remove config

* Update .gitignore

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>

* Update opendevin/runtime/docker/ssh_box.py

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>

* update setup.sh

* readd sudo

* add sudo in dockerfile

* remove export

* move od-runtime dependencies to sandbox dockerfile

* factor out re-build logic into a separate util file

* tweak existing plugin to use OD specific sandbox

* update testcase

* attempt to fix unit test using image built in ghcr

* use cache tag

* try to fix unit tests

* add unittest

* add unittest

* add some unittests

* revert gh workflow changes

* feat: optimize sandbox image naming rule

* add pull latest image hint

* add opendevin python hint and use mamba to install gcc

* update docker image naming rule and fix mamba issue

* Update opendevin/runtime/docker/ssh_box.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* fix: opendevin user use correct pip

* fix lint issue

* fix custom sandbox base image

* rename test name

* add skipif

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
Co-authored-by: tobitege <tobitege@gmx.de>
2024-06-19 19:58:07 -07:00
Xingyao Wang b569ba710d docs: Add visualizer instruction for SWE-Bench (#2529)
* Update README.md for visualizer instruction

* Polish the visualization guidance (#2531)

* fix conda create error

* fix and polish the readme for visualization

* Update README.md

---------

Co-authored-by: Haofei Yu <haofeiy@cs.cmu.edu>
2024-06-19 20:41:09 +00:00
Xingyao Wang 0a0f78f2fb docs: Update link to evaluation benchmark on README.md (#2530) 2024-06-19 20:32:39 +00:00
mamoodi c4ad3b8f47 docs: Fix formatting in CONTRIBUTING (#2526)
Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
2024-06-20 03:14:46 +08:00
dependabot[bot] dda435e5e8 chore(deps-dev): bump eslint-plugin-react in /frontend (#2524)
Bumps [eslint-plugin-react](https://github.com/jsx-eslint/eslint-plugin-react) from 7.34.2 to 7.34.3.
- [Release notes](https://github.com/jsx-eslint/eslint-plugin-react/releases)
- [Changelog](https://github.com/jsx-eslint/eslint-plugin-react/blob/master/CHANGELOG.md)
- [Commits](https://github.com/jsx-eslint/eslint-plugin-react/compare/v7.34.2...v7.34.3)

---
updated-dependencies:
- dependency-name: eslint-plugin-react
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com>
2024-06-19 17:10:08 +00:00
Graham Neubig 8ee5cbc786 Update CONTRIBUTING.md (#2525)
This PR tries to update contributing.md to add a little more structure and make it clear the different ways to contribute.
2024-06-19 15:41:48 +00:00
dependabot[bot] 64fe3d4cba chore(deps): bump framer-motion from 11.2.10 to 11.2.11 in /frontend (#2523)
Bumps [framer-motion](https://github.com/framer/motion) from 11.2.10 to 11.2.11.
- [Changelog](https://github.com/framer/motion/blob/main/CHANGELOG.md)
- [Commits](https://github.com/framer/motion/compare/v11.2.10...v11.2.11)

---
updated-dependencies:
- dependency-name: framer-motion
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-19 09:33:33 -06:00
dependabot[bot] 718e435bb8 chore(deps): bump litellm from 1.40.16 to 1.40.17 (#2522)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.16 to 1.40.17.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.16...v1.40.17)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-19 15:13:40 +00:00
dependabot[bot] 13bd154b50 chore(deps): bump boto3 from 1.34.128 to 1.34.129 (#2520)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.128 to 1.34.129.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.128...1.34.129)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-19 15:12:56 +00:00
dependabot[bot] a67b79d9ea chore(deps): bump json-repair from 0.23.1 to 0.25.0 (#2521)
Bumps [json-repair](https://github.com/mangiucugna/json_repair) from 0.23.1 to 0.25.0.
- [Release notes](https://github.com/mangiucugna/json_repair/releases)
- [Commits](https://github.com/mangiucugna/json_repair/compare/0.23.1...0.25.0)

---
updated-dependencies:
- dependency-name: json-repair
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-19 15:10:53 +00:00
dependabot[bot] e1c6af865b chore(deps-dev): bump chromadb from 0.5.1 to 0.5.3 (#2514)
Bumps [chromadb](https://github.com/chroma-core/chroma) from 0.5.1 to 0.5.3.
- [Release notes](https://github.com/chroma-core/chroma/releases)
- [Changelog](https://github.com/chroma-core/chroma/blob/main/RELEASE_PROCESS.md)
- [Commits](https://github.com/chroma-core/chroma/compare/0.5.1...0.5.3)

---
updated-dependencies:
- dependency-name: chromadb
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-19 16:49:00 +02:00
மனோஜ்குமார் பழனிச்சாமி d0bdae232f Show relevant error in UI (#2516) 2024-06-19 15:58:48 +05:30
மனோஜ்குமார் பழனிச்சாமி 92a1eaa866 Add dev config for frontend for WSL (#2506)
* Add dev config for frontend

* Update package.json

* Update package.json
2024-06-19 14:06:06 +05:30
dependabot[bot] 3ffcc7157d chore(deps-dev): bump @typescript-eslint/parser in /frontend (#2497)
Bumps [@typescript-eslint/parser](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/parser) from 7.13.0 to 7.13.1.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/parser/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.13.1/packages/parser)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/parser"
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-19 00:26:20 -07:00
Boxuan Li 89cae8d6b8 Fix Docker tagging issue with upper case (#2512)
* Fix Docker tagging issue with upper case

* Update containers/build.sh

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>

* Use tr command which is available on both zsh and bash

* Lower image name

* Lower image name

* Update .github/workflows/ghcr.yml

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>

* Fix shell syntax

---------

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>
2024-06-19 00:11:18 -06:00
tobitege 3f869ae853 default timezone in build (#2513) 2024-06-19 08:07:57 +02:00
tobitege 8f48d10442 poetry.lock updated (lots of changes) (#2492)
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-19 04:10:41 +00:00
Engel Nyst 80fe13f4be rename our completion as a drop-in replacement of litellm completion (#2509) 2024-06-19 05:25:25 +02:00
Engel Nyst b2307db010 Document, rename Agent* exceptions to LLM* (#2508)
* rename "Agent" exceptions to LLM*, document

* LLMResponseError
2024-06-18 22:30:22 +00:00
dependabot[bot] 3d33bc6813 chore(deps): bump boto3 from 1.34.126 to 1.34.128 (#2504)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.126 to 1.34.128.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.126...1.34.128)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-19 01:54:15 +08:00
Boxuan Li c2fa99b4a1 Split container image build & push (#2456)
* Split container image build & push

* Code cleanup

* Cleanup

* Add back useless docker_build_success step to make CI happy

* Revert "Cleanup"

This reverts commit 2a260791a9.

* Use fresh built sandbox image in integration test

* fix dependency

* DEBUG: only build

* Attempt to fix dependency

* Change dependency

* Combine both jobs

* Fix env

* Remove Mac integration tests as they are too unstable

* Move sandbox tests to ghcr

* Use loaded image
2024-06-19 01:44:51 +08:00
dependabot[bot] 5e8ef23a53 chore(deps): bump litellm from 1.40.15 to 1.40.16 (#2501) 2024-06-19 00:40:46 +08:00
Xingyao Wang 1f379bebc2 Update README.md (#2505)
LGTM
2024-06-18 18:14:21 +02:00
dependabot[bot] 1f72f4c401 chore(deps): bump google-generativeai from 0.6.0 to 0.7.0 (#2500) 2024-06-18 15:16:17 +00:00
dependabot[bot] 07fa57ff69 chore(deps-dev): bump @types/node from 20.14.2 to 20.14.5 in /frontend (#2498) 2024-06-18 22:54:46 +08:00
dependabot[bot] 9b9dae9463 chore(deps-dev): bump @typescript-eslint/eslint-plugin in /frontend (#2495) 2024-06-18 22:53:51 +08:00
dependabot[bot] 1bc3ca20da chore(deps): bump jose from 5.4.0 to 5.4.1 in /frontend (#2496) 2024-06-18 22:53:06 +08:00
Shimada666 e4d78153fe feat: add environments in global bashrc file (#2486) 2024-06-18 02:36:35 +00:00
dependabot[bot] e001c6a9b8 chore(deps-dev): bump ruff from 0.4.8 to 0.4.9 (#2482) 2024-06-17 16:40:50 +00:00
dependabot[bot] f8607e0ccf chore(deps): bump @nextui-org/react from 2.4.1 to 2.4.2 in /frontend (#2485) 2024-06-18 00:10:13 +08:00
dependabot[bot] f91fb3a839 chore(deps): bump tenacity from 8.3.0 to 8.4.1 (#2483) 2024-06-18 00:09:20 +08:00
dependabot[bot] 1075f5abad chore(deps): bump litellm from 1.40.12 to 1.40.15 (#2484) 2024-06-18 00:08:52 +08:00
dependabot[bot] 75ab295aec chore(deps-dev): bump flake8 from 7.0.0 to 7.1.0 (#2481) 2024-06-18 00:08:18 +08:00
மனோஜ்குமார் பழனிச்சாமி 5ae647f5c2 Bump docker version (#2479) 2024-06-17 13:50:48 +00:00
Boxuan Li 009f6f9ebc Integration tests: check agent error and fix test_edits (#2473)
* Integration tests: check agent error and fix test_edits

* Fix CodeActSWEAgent test_ipython_module prompt logs
2024-06-17 20:39:03 +08:00
மனோஜ்குமார் பழனிச்சாமி f4c917345f Reworded port forward msg (#2478)
* Reworded port forward msg

* applied suggestions.

---------

Co-authored-by: tobitege <tobitege@gmx.de>
2024-06-17 11:40:37 +00:00
tobitege 3142764104 fix: test_ipython being skipped (#2477) 2024-06-17 16:33:37 +05:30
tobitege d2509a19c8 fix: logger with more masking of sensitive data (#2470)
* fix: more logger sensitive masking

* fix: test_config.py updated for more sensitive patterns

* added one more...
2024-06-16 17:32:26 -04:00
Graham Neubig 798921cec3 Codecov after_n_builds=5 (#2468)
* Codecov after_n_builds=2

* Update to 5
2024-06-17 04:40:12 +08:00
tobitege 823298e0d0 fix: Agentskills enhancements (#2384)
* avoid repeat logging of unneeded messages

* refactored append/edit_file (tests next)

* agentskills and unit test fixes

* testing

* more changes and test prompts

* smaller changes

* final test fixes

* remove dead code from test_agent.py

* reverting unneeded changes

* updated tests, more tweaks to skills

* refactor (#2442)

* chores: fix DelegatorAgent description (#2446)

* change

* change comments

* fix

* stopped container to prevent port issues. (#2447)

* chore: remove useless browsing code in CodeActSWEAgent (#2438)

* remove useless

* fix integration test

* Regenerate test_ipython_module artifacts for CodeActSWEAgent

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Merge remote-tracking branch 'upstream/main' into agent-fileops

* unneeded tweak

* * fix edit_file to not introduce extra newline
* updated docstrings with more details for LLM
* fix legacy typo in prompts causing ]] instead of ]
* several mock files regenerated

* Regen'ed CodeActSWEAgent integration tests

* fix _print_window signature; explicit exception type in _is_valid_path

* splitlines with named param

---------

Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-16 15:06:46 -04:00
மனோஜ்குமார் பழனிச்சாமி cb44c116cd detailed jupyter error log (#2448) 2024-06-16 14:28:24 -04:00
Yufan Song 85c0eae31d remove (#2465) 2024-06-17 00:45:48 +08:00
mamoodi b5e6ba8239 docs: Update Development and CONTRIBUTING docs (#2453)
* docs: Update Development and CONTRIBUTING docs

* Explain the PR process in simpler terms

* Fix formatting

---------

Co-authored-by: Mahmoud Work <mahmoudwork@mahmouds-mini.home>
2024-06-16 10:45:10 -04:00
tobitege 2d824947f8 fix: improve toml parsing exception (#2459) 2024-06-16 14:27:21 +00:00
Boxuan Li 6f235937cf Evaluation time travel: allow evaluation on a specific version (#2356)
* Time travel for evaluation

* Fix source script path

* Exit script if given version doesn't exist

* Exit on failure

* Update README

* Change scripts of all other benchmarks

* Modify README files

* Fix logic_reasoning README
2024-06-16 10:25:14 -04:00
மனோஜ்குமார் பழனிச்சாமி 5e61a10627 Fixed typo in PR template (#2461) 2024-06-16 13:46:06 +00:00
மனோஜ்குமார் பழனிச்சாமி b5b81d06a1 Added Pull Request Template (#2454) 2024-06-16 05:54:44 -04:00
Yufan Song 426e429b18 chore: remove useless browsing code in CodeActSWEAgent (#2438)
* remove useless

* fix integration test

* Regenerate test_ipython_module artifacts for CodeActSWEAgent

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-15 10:53:03 +08:00
மனோஜ்குமார் பழனிச்சாமி ac6ef8e59d stopped container to prevent port issues. (#2447) 2024-06-15 00:31:33 +00:00
Yufan Song c82666fea3 chores: fix DelegatorAgent description (#2446)
* change

* change comments

* fix
2024-06-15 00:04:43 +00:00
Yufan Song fd29b8faa8 refactor (#2442) 2024-06-14 19:02:25 -04:00
dependabot[bot] 0d5dc1faeb chore(deps): bump vite from 5.3.0 to 5.3.1 in /frontend (#2441)
Bumps [vite](https://github.com/vitejs/vite/tree/HEAD/packages/vite) from 5.3.0 to 5.3.1.
- [Release notes](https://github.com/vitejs/vite/releases)
- [Changelog](https://github.com/vitejs/vite/blob/main/packages/vite/CHANGELOG.md)
- [Commits](https://github.com/vitejs/vite/commits/v5.3.1/packages/vite)

---
updated-dependencies:
- dependency-name: vite
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-14 23:11:35 +08:00
dependabot[bot] c272e2e851 chore(deps): bump boto3 from 1.34.125 to 1.34.126 (#2439)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.125 to 1.34.126.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.125...1.34.126)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-14 22:57:58 +08:00
dependabot[bot] 1748d47437 chore(deps): bump litellm from 1.40.9 to 1.40.12 (#2440)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.9 to 1.40.12.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.9...v1.40.12)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-14 22:57:31 +08:00
Engel Nyst bb4ea1e6cb Adjust is-stuck check for the same steps to 3 until it's stopped (#2437) 2024-06-14 19:20:12 +05:30
Boxuan Li dd1095cf6b regenerate.sh: Exit upon common known errors (#2385)
* Exit regenerate.sh upon common known errors

* More fixes

* Remove mention of transient issue

* Use tmp file instead of tty

* Remove redundant cleanup
2024-06-13 23:42:58 -07:00
Engel Nyst 1cc70be616 workspace_mount_path sentinel: an undefined string (#2431) 2024-06-14 10:39:33 +05:30
Graham Neubig 48a2f24bf4 Replace all instances of OPENDEVIN_WORKSPACE with WORKSPACE_BASE (#2418)
* Replace all instances of OPENDEVIN_WORKSPACE with WORKSPACE_BASE

* Update frontend/src/i18n/translation.json

* Update frontend/src/i18n/translation.json

---------

Co-authored-by: OpenDevin <opendevin@opendevin.ai>
Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
2024-06-13 18:27:56 +00:00
dependabot[bot] b2dc177562 chore(deps): bump datasets from 2.19.2 to 2.20.0 (#2424)
Bumps [datasets](https://github.com/huggingface/datasets) from 2.19.2 to 2.20.0.
- [Release notes](https://github.com/huggingface/datasets/releases)
- [Commits](https://github.com/huggingface/datasets/compare/2.19.2...2.20.0)

---
updated-dependencies:
- dependency-name: datasets
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-13 18:18:46 +02:00
dependabot[bot] ce0034eeaa chore(deps): bump boto3 from 1.34.124 to 1.34.125 (#2423)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.124 to 1.34.125.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.124...1.34.125)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-13 18:09:17 +02:00
dependabot[bot] d42e02f286 chore(deps-dev): bump openai from 1.33.0 to 1.34.0 (#2422)
Bumps [openai](https://github.com/openai/openai-python) from 1.33.0 to 1.34.0.
- [Release notes](https://github.com/openai/openai-python/releases)
- [Changelog](https://github.com/openai/openai-python/blob/main/CHANGELOG.md)
- [Commits](https://github.com/openai/openai-python/compare/v1.33.0...v1.34.0)

---
updated-dependencies:
- dependency-name: openai
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-13 17:49:25 +02:00
dependabot[bot] 0f26ff214a chore(deps): bump vite from 5.2.13 to 5.3.0 in /frontend (#2416)
Bumps [vite](https://github.com/vitejs/vite/tree/HEAD/packages/vite) from 5.2.13 to 5.3.0.
- [Release notes](https://github.com/vitejs/vite/releases)
- [Changelog](https://github.com/vitejs/vite/blob/main/packages/vite/CHANGELOG.md)
- [Commits](https://github.com/vitejs/vite/commits/v5.3.0/packages/vite)

---
updated-dependencies:
- dependency-name: vite
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-13 14:57:37 +00:00
dependabot[bot] 52719b5ded chore(deps-dev): bump lint-staged from 15.2.6 to 15.2.7 in /frontend (#2414)
Bumps [lint-staged](https://github.com/okonet/lint-staged) from 15.2.6 to 15.2.7.
- [Release notes](https://github.com/okonet/lint-staged/releases)
- [Changelog](https://github.com/lint-staged/lint-staged/blob/master/CHANGELOG.md)
- [Commits](https://github.com/okonet/lint-staged/compare/v15.2.6...v15.2.7)

---
updated-dependencies:
- dependency-name: lint-staged
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-13 14:56:16 +00:00
Yufan Song e04bde851e test (#2409)
Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
2024-06-13 21:10:17 +08:00
Yufan Song 7f8c324d3a Refactor CodeActSWEAgent, add response parser (#2368)
* refactor code

* fix

* seperate code

* Update agenthub/codeact_swe_agent/response_parser.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update agenthub/codeact_swe_agent/response_parser.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update agenthub/codeact_swe_agent/action_parser.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* remove browsing

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-13 04:14:05 +00:00
Yufan Song 0c92144220 Refactor MonologueAgent, PlannerAgent add response parser (#2400)
* refactor monologue

* refactor planner_agent

* fix bug

* add back code

* add back code
2024-06-13 12:00:27 +08:00
super-dainiu 563bc41fd3 Use LLM to analyze ML-Bench failure cases (#2399)
* add ml-bench w/o exec env

* fix typos (#1956)

no functional change

* Refactored Logs (#1939)

* [Feat] A competitive Web Browsing agent (#1856)

* initial attempt at a browsing only agent

* add browsing agent

* update

* implement agent

* update

* fix comments

* remove unnecessary things from memory extras

* update image processing

---------

Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>

* Update README.md SWE-bench score (#1959)

* Update README.md SWE-bench score

Our most recent results on swe-bench lite are 25%, so this updates the README accordingly.

* Update

* fix: llm is_local function logic error (#1961)

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>

* doc: update documentation about poetry update (#1962)

* add doc

* Update Development.md

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* feat: add metrics related to cost for better observability (#1944)

* add metrics for total_cost

* make lint

* refact codeact

* change metrics into llm

* add costs list, add into state

* refactor log completion

* refactor and test others

* make lint

* Update opendevin/core/metrics.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update opendevin/llm/llm.py

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>

* refactor

* add code

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>

* doc: add more cmd in unit test documentation (#1963)

* --- (#1975)

updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* --- (#1976)

updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Logging security (#1943)

* update .gitignore

* Rename the confusing 'INFO' style to 'DETAIL'

* override str and repr

* feat: api_key desensitize

* feat: add SensitiveDataFilter in file handler

* tweak regex, add tests

* more tweaks, include other attrs

* add env vars, those with equivalent config

* fix tests

* tests are invaluable

---------

Co-authored-by: Shimada666 <649940882@qq.com>

* --- (#1967)

updated-dependencies:
- dependency-name: react-dom
  dependency-type: direct:production
  update-type: version-update:semver-minor
- dependency-name: "@types/react-dom"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* --- (#1968)

updated-dependencies:
- dependency-name: "@reduxjs/toolkit"
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* --- (#1969)

updated-dependencies:
- dependency-name: husky
  dependency-type: direct:development
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* --- (#1970)

updated-dependencies:
- dependency-name: tailwind-merge
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* --- (#1971)

updated-dependencies:
- dependency-name: i18next
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>

* Refactor session management (#1810)

* refactor session mgmt

* defer file handling to runtime

* add todo

* refactor sessions a bit more

* remove messages logic from FE

* fix up socket handshake

* refactor frontend auth a bit

* first pass at redoing file explorer

* implement directory suffix

* fix up file tree

* close agent on websocket close

* remove session saving

* move file refresh

* remove getWorkspace

* plumb path/code differently

* fix build issues

* fix the tests

* fix npm build

* add session rehydration

* fix event serialization

* logspam

* fix user message rehydration

* add get_event fn

* agent state restoration

* change history tracking for codeact

* fix responsiveness of init

* fix lint

* lint

* delint

* fix prop

* update tests

* logspam

* lint

* fix test

* revert codeact

* change fileService to use API

* fix up session loading

* delint

* delint

* fix integration tests

* revert test

* fix up access to options endpoints

* fix initial files load

* delint

* fix file initialization

* fix mock server

* fixl int

* fix auth for html

* Update frontend/src/i18n/translation.json

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>

* refactor sessions and sockets

* avoid reinitializing the same session

* fix reconnect issue

* change up intro message

* more guards on reinit

* rename agent_session

* delint

* fix a bunch of tests

* delint

* fix last test

* remove code editor context

* fix build

* fix any

* fix dot notation

* Update frontend/src/services/api.ts

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* fix up error handling

* Update opendevin/server/session/agent.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update opendevin/server/session/agent.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update frontend/src/services/session.ts

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* fix build errs

* fix else

* add closed state

* delint

* Update opendevin/server/session/session.py

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>

---------

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
Co-authored-by: Graham Neubig <neubig@gmail.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>

* fix #1960 (#1964)

* Add ruff for shared mutable defaults (B) (#1938)

* Add ruff for shared mutable defaults (B)

* Apply B006, B008 on current files, except fast API

* Update agenthub/SWE_agent/prompts.py

Co-authored-by: Graham Neubig <neubig@gmail.com>

* fix unintended behavior change

* this is correct, tell Ruff to leave it alone

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Refactor integration testing CI, add optional Mac tests, and mark a few agents as deprecated (#1888)

* Add MacOS to integration tests

* Switch back to python 3.11

* Install Docker for macos pipeline

* regenerate.sh: Use environmental variable for sandbox type

* Pack different agents' tests into a single check

* Fix CodeAct tests

* Reduce file match and extensive debug logs

* Add TEST_IN_CI mode that reports codecov

* Small fix: don't quit if reusing old responses failed

* Merge codecov results

* Fix typos

* Remove coverage merge step - codecov automatically does that

* Make mac integration tests as optional - too slow

* Fix codecov args

* Add comments in yaml

* Include sandbox type in codecov report name

* Fix codecov report merge

* Revert renaming of test_matrix_success

* Remove SWEAgent and PlannerAgent from tests

* Mark planner agent and SWE agent as deprecated

* CodeCov: Ignore planner and sweagent

* Revert "Remove SWEAgent and PlannerAgent from tests"

This reverts commit 040cb3bfb9.

* Remove all tests for SWE Agent

* Only keep basic tests for MonologueAgent and PlannerAgent

* Mark SWE Agent as deprecated, and ignore code coverage for it

---------

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>

* Fix Repeated Responses in Chat by Adding IPythonRunCellObservation (#1987)

Co-authored-by: jianghongwei <jianghongwei@58.com>
Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>

* Save CI cycles for backend tests (#1985)

* Fix typo in prompt (#1992)

* Refactor monologue and SWE agent to use the messages in state history (#1863)

* Refactor monologue to use the messages in state history

* add messages, clean up

* fix monologue

* update integration tests

* move private method

* update SWE agent to use the history from State

* integration tests for SWE agent

* rename monologue to initial_thoughts, since that is what it is

* fix: catch session file not existed exception when init EventStream(maybe creating a new session with no session files stored). (#1994)

* add ml-bench in readme

* Bump boto3 from 1.34.110 to 1.34.111 (#2001)

Bumps [boto3](https://github.com/boto/boto3) from 1.34.110 to 1.34.111.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.110...1.34.111)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Bump docker from 7.0.0 to 7.1.0 (#2002)

Bumps [docker](https://github.com/docker/docker-py) from 7.0.0 to 7.1.0.
- [Release notes](https://github.com/docker/docker-py/releases)
- [Commits](https://github.com/docker/docker-py/compare/7.0.0...7.1.0)

---
updated-dependencies:
- dependency-name: docker
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Bump litellm from 1.37.20 to 1.38.0 (#2005)

Bumps [litellm](https://github.com/BerriAI/litellm) from 1.37.20 to 1.38.0.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.37.20...v1.38.0)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Fix SWE-Bench evaluation due to setuptools version (#1995)

* correctly setup plugins for swebench eval

* bump swe-bench version and add logging

* Revert "correctly setup plugins for swebench eval"

This reverts commit 2bd1055673.

* bump version

* fix session state after resuming (#1999)

* fix state resuming

* fix session reconnection

* fix lint

* Implement `agentskills` for OpenDevin to helpfully improve edit AND including more useful tools/skills (#1941)

* add draft for skills

* Implement and test agentskills functions: open_file, goto_line, scroll_down, scroll_up, create_file, search_dir, search_file, find_file

* Remove new_sample.txt file

* add some work from opendevin w/ fixes

* Add unit tests for agentskills module

* fix some issues and updated tests

* add more tests for open

* tweak and handle goto_line

* add tests for some edge cases

* add tests for scrolling

* add tests for edit

* add tests for search_dir

* update tests to use pytest

* use pytest --forked to avoid file op unit tests to interfere with each other via global var

* update doc based on swe agent tool

* update and add tests for find_file and search_file

* move agent_skills to plugins

* add agentskills as plugin and docs

* add agentskill to ssh box and fix sandbox integration

* remove extra returns in doc

* add agentskills to initial tool for jupyter

* support re-init jupyter kernel (for agentskills) after restart

* fix print window's issue with indentation and add testcases

* add prompt for codeact with the newest edit primitives

* modify the way line number is presented (remove leading space)

* change prompt to the newest display format

* support tracking of costs via metrics

* Update opendevin/runtime/plugins/agent_skills/README.md

* Update opendevin/runtime/plugins/agent_skills/README.md

* implement and add tests for py linting

* remove extra text arg for incompatible subprocess ver

* remove sample.txt

* update test_edits integration tests

* fix all integration

* Update opendevin/runtime/plugins/agent_skills/README.md

* Update opendevin/runtime/plugins/agent_skills/README.md

* Update opendevin/runtime/plugins/agent_skills/README.md

* Update agenthub/codeact_agent/prompt.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update agenthub/codeact_agent/prompt.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update agenthub/codeact_agent/prompt.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update opendevin/runtime/plugins/agent_skills/agentskills.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* correctly setup plugins for swebench eval

* bump swe-bench version and add logging

* correctly setup plugins for swebench eval

* bump swe-bench version and add logging

* Revert "correctly setup plugins for swebench eval"

This reverts commit 2bd1055673.

* bump version

* remove _AGENT_SKILLS_DOCS

* move flake8 to test dep

* update poetry.lock

* remove extra arg

* reduce max iter for eval

* update poetry

* fix integration tests

---------

Co-authored-by: OpenDevin <opendevin@opendevin.ai>
Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* build: Add poetry command to use Python 3.11 for environment setup (#1972)

* Bump @react-types/shared from 3.23.0 to 3.23.1 in /frontend (#2006)

Bumps [@react-types/shared](https://github.com/adobe/react-spectrum) from 3.23.0 to 3.23.1.
- [Release notes](https://github.com/adobe/react-spectrum/releases)
- [Commits](https://github.com/adobe/react-spectrum/compare/@react-types/shared@3.23.0...@react-types/shared@3.23.1)

---
updated-dependencies:
- dependency-name: "@react-types/shared"
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Bump @types/react-syntax-highlighter in /frontend (#2007)

Bumps [@types/react-syntax-highlighter](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/react-syntax-highlighter) from 15.5.11 to 15.5.13.
- [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases)
- [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/react-syntax-highlighter)

---
updated-dependencies:
- dependency-name: "@types/react-syntax-highlighter"
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Bump @typescript-eslint/parser from 7.9.0 to 7.10.0 in /frontend (#2008)

Bumps [@typescript-eslint/parser](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/parser) from 7.9.0 to 7.10.0.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/parser/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.10.0/packages/parser)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/parser"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Bump lint-staged from 15.2.2 to 15.2.4 in /frontend (#2009)

Bumps [lint-staged](https://github.com/okonet/lint-staged) from 15.2.2 to 15.2.4.
- [Release notes](https://github.com/okonet/lint-staged/releases)
- [Changelog](https://github.com/lint-staged/lint-staged/blob/master/CHANGELOG.md)
- [Commits](https://github.com/okonet/lint-staged/compare/v15.2.2...v15.2.4)

---
updated-dependencies:
- dependency-name: lint-staged
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Update README.md

* Update README.md

* add run_infer.sh

* fix input output

* fix docker sandbox

* fix run

* update and clean run_infer.py

* add script to clean up dockers

* update repo uid

* add description

* new

* Update README.md

* use root for sandbox

* update readme

* update ml-bench conda env

* update readme

* update readme

* use try except

* modify raise exception

* add int

* update README

* longer time

* fix existing issues

* fix existing issue

* new docker image

* add metrics of cost

* add result parsing cost

* fix

* fix

* update summarize

* fix

* add analyze

* update readme

* use 4o

* add eval output

---------

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: Ubuntu <ubuntu@ip-172-31-31-157.ec2.internal>
Co-authored-by: RainRat <rainrat78@yahoo.ca>
Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>
Co-authored-by: Frank Xu <frankxu2004@gmail.com>
Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
Co-authored-by: Graham Neubig <neubig@gmail.com>
Co-authored-by: Shimada666 <649940882@qq.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
Co-authored-by: Robert Brennan <accounts@rbren.io>
Co-authored-by: Rahul Anand <62982824+zeul22@users.noreply.github.com>
Co-authored-by: jiangleo <jiangleo@users.noreply.github.com>
Co-authored-by: jianghongwei <jianghongwei@58.com>
Co-authored-by: Jeremi Joslin <jeremi@newlogic.com>
Co-authored-by: Aaron Xia <zhhuaxia@gmail.com>
Co-authored-by: OpenDevin <opendevin@opendevin.ai>
Co-authored-by: DaxServer <7479937+DaxServer@users.noreply.github.com>
Co-authored-by: Robert <871607149@qq.com>
2024-06-13 09:30:55 +08:00
dependabot[bot] 0bd5ab72ed chore(deps): bump boto3 from 1.34.123 to 1.34.124 (#2410) 2024-06-13 00:07:45 +08:00
dependabot[bot] 4c774544ce chore(deps): bump litellm from 1.40.8 to 1.40.9 (#2411) 2024-06-13 00:07:09 +08:00
dependabot[bot] d1eaf486bd chore(deps-dev): bump lint-staged from 15.2.5 to 15.2.6 in /frontend (#2407)
Bumps [lint-staged](https://github.com/okonet/lint-staged) from 15.2.5 to 15.2.6.
- [Release notes](https://github.com/okonet/lint-staged/releases)
- [Changelog](https://github.com/lint-staged/lint-staged/blob/master/CHANGELOG.md)
- [Commits](https://github.com/okonet/lint-staged/compare/v15.2.5...v15.2.6)

---
updated-dependencies:
- dependency-name: lint-staged
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-12 17:35:01 +03:00
Leo 7fcd14a055 fix the failed unit test. (#2405)
Signed-off-by: ifuryst <ifuryst@gmail.com>
2024-06-12 13:55:23 +00:00
Yufan Song 90ec0095df Add integration test for CodeActSWEAgent (#2377)
* add test log

* remove browsing internet

* add test by GPT-4o

* fix prompts

* change test_agent

* fix test

* fix nits
2024-06-12 02:46:15 +08:00
dependabot[bot] a25fb4d21b chore(deps-dev): bump @typescript-eslint/parser in /frontend (#2387)
Bumps [@typescript-eslint/parser](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/parser) from 7.12.0 to 7.13.0.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/parser/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.13.0/packages/parser)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/parser"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-11 18:33:20 +00:00
dependabot[bot] 11d7efb0e8 chore(deps-dev): bump @testing-library/jest-dom in /frontend (#2388)
Bumps [@testing-library/jest-dom](https://github.com/testing-library/jest-dom) from 6.4.5 to 6.4.6.
- [Release notes](https://github.com/testing-library/jest-dom/releases)
- [Changelog](https://github.com/testing-library/jest-dom/blob/main/CHANGELOG.md)
- [Commits](https://github.com/testing-library/jest-dom/compare/v6.4.5...v6.4.6)

---
updated-dependencies:
- dependency-name: "@testing-library/jest-dom"
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-12 02:17:20 +08:00
dependabot[bot] 17e7a45c3c chore(deps-dev): bump @typescript-eslint/eslint-plugin in /frontend (#2389)
Bumps [@typescript-eslint/eslint-plugin](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/eslint-plugin) from 7.12.0 to 7.13.0.
- [Release notes](https://github.com/typescript-eslint/typescript-eslint/releases)
- [Changelog](https://github.com/typescript-eslint/typescript-eslint/blob/main/packages/eslint-plugin/CHANGELOG.md)
- [Commits](https://github.com/typescript-eslint/typescript-eslint/commits/v7.13.0/packages/eslint-plugin)

---
updated-dependencies:
- dependency-name: "@typescript-eslint/eslint-plugin"
  dependency-type: direct:development
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-12 02:16:47 +08:00
dependabot[bot] e14297fde9 chore(deps-dev): bump prettier from 3.3.1 to 3.3.2 in /frontend (#2390)
Bumps [prettier](https://github.com/prettier/prettier) from 3.3.1 to 3.3.2.
- [Release notes](https://github.com/prettier/prettier/releases)
- [Changelog](https://github.com/prettier/prettier/blob/main/CHANGELOG.md)
- [Commits](https://github.com/prettier/prettier/compare/3.3.1...3.3.2)

---
updated-dependencies:
- dependency-name: prettier
  dependency-type: direct:development
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-12 02:15:00 +08:00
dependabot[bot] 35c4c9cb34 chore(deps): bump litellm from 1.40.7 to 1.40.8 (#2392)
Bumps [litellm](https://github.com/BerriAI/litellm) from 1.40.7 to 1.40.8.
- [Release notes](https://github.com/BerriAI/litellm/releases)
- [Commits](https://github.com/BerriAI/litellm/compare/v1.40.7...v1.40.8)

---
updated-dependencies:
- dependency-name: litellm
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-11 15:23:14 +00:00
dependabot[bot] 9893c06b2e chore(deps): bump boto3 from 1.34.122 to 1.34.123 (#2391)
Bumps [boto3](https://github.com/boto/boto3) from 1.34.122 to 1.34.123.
- [Release notes](https://github.com/boto/boto3/releases)
- [Changelog](https://github.com/boto/boto3/blob/develop/CHANGELOG.rst)
- [Commits](https://github.com/boto/boto3/compare/1.34.122...1.34.123)

---
updated-dependencies:
- dependency-name: boto3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2024-06-11 15:19:34 +00:00
Yufan Song c6951eb6c1 refactor browsing agent response parse (#2366)
* refactor browsing

* add comments

* change file name

* Rename resposne_parser.py to response_parser.py

* Fixed typos

* Typo fix

---------

Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-11 03:46:33 +00:00
Xingyao Wang b3bdc44292 mkdir infer_logs instead of logs (#2382) 2024-06-11 07:18:19 +08:00
482 changed files with 34542 additions and 14858 deletions
+3 -3
View File
@@ -1,6 +1,9 @@
codecov:
notify:
wait_for_ci: true
# our project is large, so 6 builds are typically uploaded. this waits till 5/6
# See https://docs.codecov.com/docs/notifications#section-preventing-notifications-until-after-n-builds
after_n_builds: 5
coverage:
status:
@@ -14,6 +17,3 @@ coverage:
comment: false
github_checks:
annotations: false
ignore:
- "agenthub/SWE_agent/**" # SWE agent is deprecated
+1 -1
View File
@@ -72,4 +72,4 @@ body:
id: additional-context
attributes:
label: Logs, Errors, Screenshots, and Additional Context
description: LLM logs will be stored in the `logs/llm/default` folder. Please add any additional context about the problem here.
description: If you want to share the chat history you can click the thumbs-down (👎) button above the input field and you will get a shareable link (you can also click thumbs up when things are going well of course!). LLM logs will be stored in the `logs/llm/default` folder. Please add any additional context about the problem here.
+7
View File
@@ -9,7 +9,14 @@ updates:
directory: "/" # Location of package manifests
schedule:
interval: "daily"
open-pull-requests-limit: 20
- package-ecosystem: "npm" # See documentation for possible values
directory: "/frontend" # Location of package manifests
schedule:
interval: "daily"
open-pull-requests-limit: 20
- package-ecosystem: "npm" # See documentation for possible values
directory: "/docs" # Location of package manifests
schedule:
interval: "daily"
open-pull-requests-limit: 20
+5
View File
@@ -0,0 +1,5 @@
**What is the problem that this fixes or functionality that this introduces? Does it fix any open issues?**
**Give a brief summary of what the PR does, explaining any non-trivial design decisions**
**Other references**
+11 -2
View File
@@ -1,4 +1,4 @@
name: Run e2e test with dummy agent
name: Run E2E test with dummy agent
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
@@ -30,4 +30,13 @@ jobs:
wget https://huggingface.co/BAAI/bge-small-en-v1.5/raw/main/1_Pooling/config.json -P /tmp/llama_index/models--BAAI--bge-small-en-v1.5/snapshots/5c38ec7c405ec4b44b94cc5a9bb96e735b38267a/1_Pooling/
- name: Run tests
run: |
poetry run python opendevin/core/main.py -t "do a flip" -m ollama/not-a-model -d ./workspace/ -c DummyAgent
set -e
poetry run python opendevin/core/main.py -t "do a flip" -d ./workspace/ -c DummyAgent
- name: Check exit code
run: |
if [ $? -ne 0 ]; then
echo "Test failed"
exit 1
else
echo "Test passed"
fi
+200 -19
View File
@@ -1,4 +1,4 @@
name: Publish Docker Image
name: Build Publish and Test Docker Image
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
@@ -7,7 +7,7 @@ concurrency:
on:
push:
branches:
- main
- main
tags:
- '*'
pull_request:
@@ -19,19 +19,23 @@ on:
default: ''
jobs:
ghcr_build_and_push:
ghcr_build:
runs-on: ubuntu-latest
outputs:
tags: ${{ steps.capture-tags.outputs.tags }}
permissions:
contents: read
packages: write
strategy:
matrix:
image: ["app", "sandbox"]
image: ["sandbox", "opendevin"]
platform: ["amd64", "arm64"]
steps:
- name: checkout
- name: Checkout
uses: actions/checkout@v4
- name: Free Disk Space (Ubuntu)
@@ -40,7 +44,6 @@ jobs:
# this might remove tools that are actually needed,
# if set to "true" but frees about 6 GB
tool-cache: true
# all of these default to true, but feel free to set to
# "false" if necessary for your workflow
android: true
@@ -57,26 +60,204 @@ jobs:
id: buildx
uses: docker/setup-buildx-action@v3
- name: Login to ghcr
uses: docker/login-action@v1
- name: Build and export image
id: build
run: ./containers/build.sh ${{ matrix.image }} ${{ github.repository_owner }} ${{ matrix.platform }}
- name: Capture tags
id: capture-tags
run: |
tags=$(cat tags.txt)
echo "tags=$tags"
echo "tags=$tags" >> $GITHUB_OUTPUT
- name: Upload Docker image as artifact
uses: actions/upload-artifact@v4
with:
name: ${{ matrix.image }}-docker-image-${{ matrix.platform }}
path: /tmp/${{ matrix.image }}_image_${{ matrix.platform }}.tar
test-for-sandbox:
name: Test for Sandbox
runs-on: ubuntu-latest
needs: ghcr_build
env:
PERSIST_SANDBOX: "false"
steps:
- uses: actions/checkout@v4
- name: Install poetry via pipx
run: pipx install poetry
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: "3.11"
cache: "poetry"
- name: Install Python dependencies using Poetry
run: make install-python-dependencies
- name: Download sandbox Docker image
uses: actions/download-artifact@v4
with:
name: sandbox-docker-image-amd64
path: /tmp/
- name: Load sandbox image and run sandbox tests
run: |
# Load the Docker image and capture the output
output=$(docker load -i /tmp/sandbox_image_amd64.tar)
# Extract the first image name from the output
image_name=$(echo "$output" | grep -oP 'Loaded image: \K.*' | head -n 1)
# Print the full name of the image
echo "Loaded Docker image: $image_name"
SANDBOX_CONTAINER_IMAGE=$image_name TEST_IN_CI=true poetry run pytest --cov=agenthub --cov=opendevin --cov-report=xml -s ./tests/unit/test_sandbox.py
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v4
env:
CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
integration-tests-on-linux:
name: Integration Tests on Linux
runs-on: ubuntu-latest
needs: ghcr_build
env:
PERSIST_SANDBOX: "false"
strategy:
fail-fast: false
matrix:
python-version: ["3.11"]
sandbox: ["ssh", "local"]
steps:
- uses: actions/checkout@v4
- name: Install poetry via pipx
run: pipx install poetry
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
cache: 'poetry'
- name: Install Python dependencies using Poetry
run: make install-python-dependencies
- name: Download sandbox Docker image
uses: actions/download-artifact@v4
with:
name: sandbox-docker-image-amd64
path: /tmp/
- name: Load sandbox image and run integration tests
env:
SANDBOX_BOX_TYPE: ${{ matrix.sandbox }}
run: |
# Load the Docker image and capture the output
output=$(docker load -i /tmp/sandbox_image_amd64.tar)
# Extract the first image name from the output
image_name=$(echo "$output" | grep -oP 'Loaded image: \K.*' | head -n 1)
# Print the full name of the image
echo "Loaded Docker image: $image_name"
SANDBOX_CONTAINER_IMAGE=$image_name TEST_IN_CI=true TEST_ONLY=true ./tests/integration/regenerate.sh
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v4
env:
CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
ghcr_push:
runs-on: ubuntu-latest
# don't push if integration tests or sandbox tests fail
needs: [ghcr_build, integration-tests-on-linux, test-for-sandbox]
if: github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/')
env:
tags: ${{ needs.ghcr_build.outputs.tags }}
permissions:
contents: read
packages: write
strategy:
matrix:
image: ["sandbox", "opendevin"]
platform: ["amd64", "arm64"]
steps:
- name: Checkout code
uses: actions/checkout@v4
- name: Login to GHCR
uses: docker/login-action@v2
with:
registry: ghcr.io
username: ${{ github.repository_owner }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Build and push ${{ matrix.image }}
if: "!github.event.pull_request.head.repo.fork"
run: |
./containers/build.sh ${{ matrix.image }} ${{ github.repository_owner }} --push
- name: Download Docker images
uses: actions/download-artifact@v4
with:
name: ${{ matrix.image }}-docker-image-${{ matrix.platform }}
path: /tmp/${{ matrix.platform }}
- name: Build ${{ matrix.image }}
if: "github.event.pull_request.head.repo.fork"
- name: Load images and push to registry
run: |
./containers/build.sh ${{ matrix.image }} ${{ github.repository_owner }}
mv /tmp/${{ matrix.platform }}/${{ matrix.image }}_image_${{ matrix.platform }}.tar .
loaded_image=$(docker load -i ${{ matrix.image }}_image_${{ matrix.platform }}.tar | grep "Loaded image:" | head -n 1 | awk '{print $3}')
echo "loaded image = $loaded_image"
tags=$(echo ${tags} | tr ' ' '\n')
image_name=$(echo "ghcr.io/${{ github.repository_owner }}/${{ matrix.image }}" | tr '[:upper:]' '[:lower:]')
echo "image name = $image_name"
for tag in $tags; do
echo "tag = $tag"
docker tag $loaded_image $image_name:${tag}_${{ matrix.platform }}
docker push $image_name:${tag}_${{ matrix.platform }}
done
docker_build_success:
name: Docker Build Success
create_manifest:
runs-on: ubuntu-latest
needs: ghcr_build_and_push
needs: [ghcr_build, ghcr_push]
if: github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/')
env:
tags: ${{ needs.ghcr_build.outputs.tags }}
strategy:
matrix:
image: ["sandbox", "opendevin"]
permissions:
contents: read
packages: write
steps:
- run: echo Done!
- name: Checkout code
uses: actions/checkout@v4
- name: Login to GHCR
uses: docker/login-action@v2
with:
registry: ghcr.io
username: ${{ github.repository_owner }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Create and push multi-platform manifest
run: |
image_name=$(echo "ghcr.io/${{ github.repository_owner }}/${{ matrix.image }}" | tr '[:upper:]' '[:lower:]')
echo "image name = $image_name"
tags=$(echo ${tags} | tr ' ' '\n')
for tag in $tags; do
echo 'tag = $tag'
docker buildx imagetools create --tag $image_name:$tag \
$image_name:${tag}_amd64 \
$image_name:${tag}_arm64
done
+1 -1
View File
@@ -47,4 +47,4 @@ jobs:
- name: Install pre-commit
run: pip install pre-commit==3.7.0
- name: Run pre-commit hooks
run: pre-commit run --files opendevin/**/* agenthub/**/* evaluation/**/* --show-diff-on-failure --config ./dev_config/python/.pre-commit-config.yaml
run: pre-commit run --files opendevin/**/* agenthub/**/* evaluation/**/* tests/**/* --show-diff-on-failure --config ./dev_config/python/.pre-commit-config.yaml
+1 -1
View File
@@ -55,7 +55,7 @@ jobs:
env:
LLM_API_KEY: ${{ secrets.OPENAI_API_KEY }}
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
SANDBOX_TYPE: exec
SANDBOX_BOX_TYPE: ssh
run: |
# Append path to launch poetry
export PATH="/github/home/.local/bin:$PATH"
-104
View File
@@ -1,104 +0,0 @@
name: Run Integration Tests
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: ${{ github.ref != 'refs/heads/main' }}
on:
push:
branches:
- main
paths-ignore:
- '**/*.md'
- 'frontend/**'
- 'docs/**'
- 'evaluation/**'
pull_request:
env:
PERSIST_SANDBOX : "false"
jobs:
integration-tests-on-linux:
name: Integration Tests on Linux
runs-on: ubuntu-latest
strategy:
fail-fast: false
matrix:
python-version: ["3.11"]
sandbox: ["ssh", "exec", "local"]
steps:
- uses: actions/checkout@v4
- name: Install poetry via pipx
run: pipx install poetry
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
cache: 'poetry'
- name: Install Python dependencies using Poetry
run: poetry install
- name: Build Environment
run: make build
- name: Run Integration Tests
env:
SANDBOX_TYPE: ${{ matrix.sandbox }}
run: |
TEST_IN_CI=true TEST_ONLY=true ./tests/integration/regenerate.sh
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v4
env:
CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
integration-tests-on-mac:
name: Integration Tests on MacOS
runs-on: macos-13
if: contains(github.event.pull_request.title, 'mac') || contains(github.event.pull_request.title, 'Mac')
strategy:
fail-fast: false
matrix:
python-version: ["3.11"]
sandbox: ["ssh"]
steps:
- uses: actions/checkout@v4
- name: Install poetry via pipx
run: pipx install poetry
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
cache: 'poetry'
- name: Install Python dependencies using Poetry
run: poetry install
- name: Install & Start Docker
run: |
brew install colima docker
colima start
# For testcontainers to find the Colima socket
# https://github.com/abiosoft/colima/blob/main/docs/FAQ.md#cannot-connect-to-the-docker-daemon-at-unixvarrundockersock-is-the-docker-daemon-running
sudo ln -sf $HOME/.colima/default/docker.sock /var/run/docker.sock
- name: Build Environment
run: make build
- name: Run Integration Tests
env:
SANDBOX_TYPE: ${{ matrix.sandbox }}
run: |
TEST_IN_CI=true TEST_ONLY=true ./tests/integration/regenerate.sh
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v4
env:
CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
+42 -33
View File
@@ -19,9 +19,38 @@ env:
PERSIST_SANDBOX : "false"
jobs:
fe-test:
runs-on: ubuntu-latest
strategy:
matrix:
node-version: [20]
steps:
- name: Checkout
uses: actions/checkout@v4
- name: Set up Node.js
uses: actions/setup-node@v4
with:
node-version: ${{ matrix.node-version }}
- name: Install dependencies
working-directory: ./frontend
run: npm ci
- name: Run tests and collect coverage
working-directory: ./frontend
run: npm run test:coverage
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v4
env:
CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
test-on-macos:
name: Test on macOS
runs-on: macos-13
runs-on: macos-12
env:
INSTALL_DOCKER: "1" # Set to '0' to skip Docker installation
strategy:
@@ -46,8 +75,18 @@ jobs:
- name: Install & Start Docker
if: env.INSTALL_DOCKER == '1'
run: |
brew install colima docker
colima start
# Uninstall colima to upgrade to the latest version
if brew list colima &>/dev/null; then
brew uninstall colima
# unlinking colima dependency: go
brew uninstall go@1.21
fi
rm -rf ~/.colima ~/.lima
brew install --HEAD colima
brew services start colima
brew install docker
colima delete
colima start --network-address --arch x86_64 --cpu=1 --memory=1
# For testcontainers to find the Colima socket
# https://github.com/abiosoft/colima/blob/main/docs/FAQ.md#cannot-connect-to-the-docker-daemon-at-unixvarrundockersock-is-the-docker-daemon-running
@@ -97,33 +136,3 @@ jobs:
uses: codecov/codecov-action@v4
env:
CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
test-for-sandbox:
name: Test for Sandbox
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Install poetry via pipx
run: pipx install poetry
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: "3.11"
cache: "poetry"
- name: Install Python dependencies using Poetry
run: poetry install
- name: Build Environment
run: make build
- name: Run Integration Test for Sandbox
run: |
poetry run pytest --cov=agenthub --cov=opendevin --cov-report=xml -s ./tests/unit/test_sandbox.py
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v4
env:
CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
+2 -2
View File
@@ -35,7 +35,7 @@ jobs:
echo "" >> task.txt
echo "BODY:" >> task.txt
echo "${ISSUE_BODY}" >> task.txt
- name: Set up environment
run: |
curl -sSL https://install.python-poetry.org | python3 -
@@ -50,7 +50,7 @@ jobs:
ISSUE_BODY: ${{ github.event.issue.body }}
LLM_API_KEY: ${{ secrets.OPENAI_API_KEY }}
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
SANDBOX_TYPE: exec
SANDBOX_BOX_TYPE: ssh
run: |
# Append path to launch poetry
export PATH="/github/home/.local/bin:$PATH"
@@ -0,0 +1,48 @@
name: Update pyproject.toml Version and Tags
on:
release:
types:
- published
jobs:
update-pyproject-and-tags:
runs-on: ubuntu-latest
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
fetch-depth: 0 # Fetch all history for all branches and tags
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: "3.11"
- name: Install dependencies
run: |
python -m pip install --upgrade pip
pip install toml
- name: Get release tag
id: get_release_tag
run: echo "RELEASE_TAG=${GITHUB_REF#refs/tags/}" >> $GITHUB_ENV
- name: Update pyproject.toml with release tag
run: |
python -c "
import toml
with open('pyproject.toml', 'r') as f:
data = toml.load(f)
data['tool']['poetry']['version'] = '${{ env.RELEASE_TAG }}'
with open('pyproject.toml', 'w') as f:
toml.dump(data, f)
"
- name: Commit and push pyproject.toml changes
uses: stefanzweifel/git-auto-commit-action@v4
with:
commit_message: "Update pyproject.toml version to ${{ env.RELEASE_TAG }}"
branch: main
file_pattern: pyproject.toml
+6
View File
@@ -211,3 +211,9 @@ cache
# configuration
config.toml
config.toml.bak
containers/agnostic_sandbox
# swe-bench-eval
image_build_logs
run_instance_logs
+62 -35
View File
@@ -1,13 +1,32 @@
# Contributing
Thanks for your interest in contributing to OpenDevin! We welcome and appreciate contributions.
To report bugs, create a [GitHub issue](https://github.com/OpenDevin/OpenDevin/issues/new/choose).
Thanks for your interest in contributing to OpenDevin! We welcome and appreciate contributions.
## How Can I Contribute?
There are many ways that you can contribute:
1. **Download and use** OpenDevin, and send [issues](https://github.com/OpenDevin/OpenDevin/issues) when you encounter something that isn't working or a feature that you'd like to see.
2. **Send feedback** after each session by [clicking the thumbs-up thumbs-down buttons](https://opendevin.github.io/OpenDevin/modules/usage/feedback), so we can see where things are working and failing, and also build an open dataset for training code agents.
3. **Improve the Codebase** by sending PRs (see details below). In particular, we have some [good first issue](https://github.com/OpenDevin/OpenDevin/labels/good%20first%20issue) issues that may be ones to start on.
## Understanding OpenDevin's CodeBase
To understand the codebase, please refer to the README in each module:
- [frontend](./frontend/README.md)
- [agenthub](./agenthub/README.md)
- [evaluation](./evaluation/README.md)
- [opendevin](./opendevin/README.md)
- [server](./opendevin/server/README.md)
When you write code, it is also good to write tests. Please navigate to the `tests` folder to see existing test suites.
At the moment, we have two kinds of tests: `unit` and `integration`. Please refer to the README for each test suite. These tests also run on GitHub's continuous integration to ensure quality of the project.
## Sending Pull Requests to OpenDevin
## Contribution Guide
### 1. Fork the Official Repository
Fork [OpenDevin repository](https://github.com/OpenDevin/OpenDevin) into your own account.
Clone your own forked repository into your local environment.
Fork the [OpenDevin repository](https://github.com/OpenDevin/OpenDevin) into your own account.
Clone your own forked repository into your local environment:
```shell
git clone git@github.com:<YOUR-USERNAME>/OpenDevin.git
@@ -16,21 +35,23 @@ git clone git@github.com:<YOUR-USERNAME>/OpenDevin.git
### 2. Configure Git
Set the official repository as your [upstream](https://www.atlassian.com/git/tutorials/git-forks-and-upstreams) to synchronize with the latest update in the official repository.
Add the original repository as upstream
Add the original repository as upstream:
```shell
cd OpenDevin
git remote add upstream git@github.com:OpenDevin/OpenDevin.git
```
Verify that the remote is set.
Verify that the remote is set:
```shell
git remote -v
```
You should see both `origin` and `upstream` in the output.
### 3. Synchronize with Official Repository
Synchronize latest commit with official repository before coding.
Synchronize latest commit with official repository before coding:
```shell
git fetch upstream
@@ -39,25 +60,48 @@ git merge upstream/main
git push origin main
```
### 4. Create a New Branch And Open a Pull Request
After you finish implementation, open forked repository. The source branch is your new branch, and the target branch is `OpenDevin/OpenDevin` `main` branch. Then PR should appears in [OpenDevin PRs](https://github.com/OpenDevin/OpenDevin/pulls).
### 4. Set up the Development Environment
Then OpenDevin team will review your code.
We have a separate doc [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md) that tells you how to set up a development workflow.
### 5. Write Code and Commit It
Once you have done this, you can write code, test it, and commit it to a branch (replace `my_branch` with an appropriate name):
```shell
git checkout -b my_branch
git add .
git commit
git push origin my_branch
```
### 6. Open a Pull Request
* On GitHub, go to the page of your forked repository, and create a Pull Request:
- Click on `Branches`
- Click on the `...` beside your branch and click on `New pull request`
- Set `base repository` to `OpenDevin/OpenDevin`
- Set `base` to `main`
- Click `Create pull request`
The PR should appear in [OpenDevin PRs](https://github.com/OpenDevin/OpenDevin/pulls).
Then the OpenDevin team will review your code.
## PR Rules
### 1. Pull Request title
As described in [here](https://github.com/commitizen/conventional-commit-types/blob/master/index.json), a valid PR title should begin with one of the following prefixes:
As described [here](https://github.com/commitizen/conventional-commit-types/blob/master/index.json), a valid PR title should begin with one of the following prefixes:
- `feat`: A new feature
- `fix`: A bug fix
- `doc`: Documentation only changes
- `docs`: Documentation only changes
- `style`: Changes that do not affect the meaning of the code (white space, formatting, missing semicolons, etc.)
- `refactor`: A code change that neither fixes a bug nor adds a feature
- `style`: A refactoring that improves code style
- `perf`: A code change that improves performance
- `test`: Adding missing tests or correcting existing tests
- `ci`: Changes to CI configuration files and scripts (example scopes: `.github`, `ci` (Buildkite))
- `build`: Changes that affect the build system or external dependencies (example scopes: gulp, broccoli, npm)
- `ci`: Changes to our CI configuration files and scripts (example scopes: Travis, Circle, BrowserStack, SauceLabs)
- `chore`: Other changes that don't modify src or test files
- `revert`: Reverts a previous commit
@@ -67,24 +111,7 @@ For example, a PR title could be:
You may also check out previous PRs in the [PR list](https://github.com/OpenDevin/OpenDevin/pulls).
As described in [here](https://github.com/OpenDevin/OpenDevin/labels), we create several labels. Every PR should be tagged with the corresponding labels.
### 2. Pull Request description
- If your PR is small (such as a typo fix), you can go brief.
- If it is large and you have changed a lot, it's better to write more details.
- If it contains a lot of changes, it's better to write more details.
## How to begin
Please refer to the README in each module:
- [frontend](./frontend/README.md)
- [agenthub](./agenthub/README.md)
- [evaluation](./evaluation/README.md)
- [opendevin](./opendevin/README.md)
- [server](./opendevin/server/README.md)
- [mock server](./opendevin/mock/README.md)
## Tests
Please navigate to `tests` folder to see existing test suites.
At the moment, we have two kinds of tests: `unit` and `integration`. Please refer to the README for each test suite. These tests also run on CI to ensure quality of
the project.
+29 -33
View File
@@ -1,14 +1,16 @@
# Development Guide
This guide is for people working on OpenDevin and editing the source code.
If you wish to contribute your changes, check out the [CONTRIBUTING.md](https://github.com/OpenDevin/OpenDevin/blob/main/CONTRIBUTING.md) on how to clone and setup the project initially before moving on.
Otherwise, you can clone the OpenDevin project directly.
## Start the server for development
### 1. Requirements
* Linux, Mac OS, or [WSL on Windows](https://learn.microsoft.com/en-us/windows/wsl/install)
* Linux, Mac OS, or [WSL on Windows](https://learn.microsoft.com/en-us/windows/wsl/install) [ Ubuntu <= 22.04]
* [Docker](https://docs.docker.com/engine/install/) (For those on MacOS, make sure to allow the default Docker socket to be used from advanced settings!)
* [Python](https://www.python.org/downloads/) = 3.11
* [NodeJS](https://nodejs.org/en/download/package-manager) >= 18.17.1
* [Poetry](https://python-poetry.org/docs/#installing-with-the-official-installer) >= 1.8
* netcat => sudo apt-get install netcat
Make sure you have all these dependencies installed before moving on to `make build`.
@@ -27,42 +29,40 @@ mamba install conda-forge::poetry
```
### 2. Build and Setup The Environment
Begin by building the project which includes setting up the environment and installing dependencies. This step ensures that OpenDevin is ready to run on your system:
- **Build the Project:** Begin by building the project, which includes setting up the environment and installing dependencies. This step ensures that OpenDevin is ready to run smoothly on your system.
```bash
make build
```
```bash
make build
```
### 3. Configuring the Language Model
OpenDevin supports a diverse array of Language Models (LMs) through the powerful [litellm](https://docs.litellm.ai) library. By default, we've chosen the mighty GPT-4 from OpenAI as our go-to model, but the world is your oyster! You can unleash the potential of Anthropic's suave Claude, the enigmatic Llama, or any other LM that piques your interest.
To configure the LM of your choice, follow these steps:
1. **Using the Makefile: The Effortless Approach**
With a single command, you can have a smooth LM setup for your OpenDevin experience. Simply run:
To configure the LM of your choice, run:
```bash
make setup-config
```
This command will prompt you to enter the LLM API key, model name, and other variables ensuring that OpenDevin is tailored to your specific needs. Note that the model name will apply only when you run headless. If you use the UI, please set the model in the UI.
Set `persist_sandbox` to false if you want to use clean sandbox for each task. If `persist_sandbox` is set to true, you will need to set the `ssh_password` as well.
Note: If you have previously run OpenDevin using the docker command, you may have already set some environmental variables in your terminal. The final configurations are set from highest to lowest priority:
Environment variables > config.toml variables > default variables
**Note on Alternative Models:**
Some alternative models may prove more challenging to tame than others. Fear not, brave adventurer! We shall soon unveil LLM-specific documentation to guide you on your quest. And if you've already mastered the art of wielding a model other than OpenAI's GPT, we encourage you to [share your setup instructions with us](https://github.com/OpenDevin/OpenDevin/issues/417).
Some alternative models may prove more challenging to tame than others. Fear not, brave adventurer! We shall soon unveil LLM-specific documentation to guide you on your quest.
And if you've already mastered the art of wielding a model other than OpenAI's GPT, we encourage you to share your setup instructions with us by creating instructions and adding it [to our documentation](https://github.com/OpenDevin/OpenDevin/tree/main/docs/modules/usage/llms).
For a full list of the LM providers and models available, please consult the [litellm documentation](https://docs.litellm.ai/docs/providers).
There is also [documentation for running with local models using ollama](./docs/documentation/LOCAL_LLM_GUIDE.md).
### 4. Run the Application
- **Run the Application:** Once the setup is complete, launching OpenDevin is as simple as running a single command. This command starts both the backend and frontend servers seamlessly, allowing you to interact with OpenDevin without any hassle.
```bash
make run
```
### 5. Individual Server Startup
### 4. Running the application
#### Option A: Run the Full Application
Once the setup is complete, launching OpenDevin is as simple as running a single command. This command starts both the backend and frontend servers seamlessly, allowing you to interact with OpenDevin:
```bash
make run
```
#### Option B: Individual Server Startup
- **Start the Backend Server:** If you prefer, you can start the backend server independently to focus on backend-related tasks or configurations.
```bash
make start-backend
@@ -74,18 +74,16 @@ There is also [documentation for running with local models using ollama](./docs/
```
### 6. LLM Debugging
If you encounter any issues with the Language Model (LM) or you're simply curious, you can inspect the actual LLM prompts and responses. To do so, export DEBUG=1 in the environment and restart the backend. OpenDevin will then log the prompts and responses in the logs/llm/CURRENT_DATE directory, allowing you to identify the causes.
If you encounter any issues with the Language Model (LM) or you're simply curious, you can inspect the actual LLM prompts and responses. To do so, export DEBUG=1 in the environment and restart the backend.
OpenDevin will then log the prompts and responses in the logs/llm/CURRENT_DATE directory, allowing you to identify the causes.
### 7. Help
- **Get Some Help:** Need assistance or information on available targets and commands? The help command provides all the necessary guidance to ensure a smooth experience with OpenDevin.
```bash
make help
```
Need assistance or information on available targets and commands? The help command provides all the necessary guidance to ensure a smooth experience with OpenDevin.
```bash
make help
```
### 8. Testing
#### Unit tests
```bash
@@ -93,10 +91,8 @@ poetry run pytest ./tests/unit/test_sandbox.py
```
#### Integration tests
Please refer to [this README](./tests/integration/README.md) for details.
### 9. Add or update dependency
1. Add your dependency in `pyproject.toml` or use `poetry add xxx`
2. Update the poetry.lock file via `poetry lock --no-update`
+28 -13
View File
@@ -2,14 +2,14 @@ SHELL=/bin/bash
# Makefile for OpenDevin project
# Variables
DOCKER_IMAGE = ghcr.io/opendevin/sandbox
DOCKER_IMAGE = ghcr.io/opendevin/sandbox:main
BACKEND_PORT = 3000
BACKEND_HOST = "127.0.0.1:$(BACKEND_PORT)"
FRONTEND_PORT = 3001
DEFAULT_WORKSPACE_DIR = "./workspace"
DEFAULT_MODEL = "gpt-4o"
CONFIG_FILE = config.toml
PRECOMMIT_CONFIG_PATH = "./dev_config/python/.pre-commit-config.yaml"
PRE_COMMIT_CONFIG_PATH = "./dev_config/python/.pre-commit-config.yaml"
PYTHON_VERSION = 3.11
# ANSI color codes
@@ -28,7 +28,7 @@ ifeq ($(INSTALL_DOCKER),)
endif
@$(MAKE) -s install-python-dependencies
@$(MAKE) -s install-frontend-dependencies
@$(MAKE) -s install-precommit-hooks
@$(MAKE) -s install-pre-commit-hooks
@$(MAKE) -s build-frontend
@echo "$(GREEN)Build completed successfully.$(RESET)"
@@ -131,6 +131,10 @@ pull-docker-image:
install-python-dependencies:
@echo "$(GREEN)Installing Python dependencies...$(RESET)"
@if [ -z "${TZ}" ]; then \
echo "Defaulting TZ (timezone) to UTC"; \
export TZ="UTC"; \
fi
poetry env use python$(PYTHON_VERSION)
@if [ "$(shell uname)" = "Darwin" ]; then \
echo "$(BLUE)Installing chroma-hnswlib...$(RESET)"; \
@@ -165,15 +169,15 @@ install-frontend-dependencies:
npm run make-i18n
@echo "$(GREEN)Frontend dependencies installed successfully.$(RESET)"
install-precommit-hooks:
install-pre-commit-hooks:
@echo "$(YELLOW)Installing pre-commit hooks...$(RESET)"
@git config --unset-all core.hooksPath || true
@poetry run pre-commit install --config $(PRECOMMIT_CONFIG_PATH)
@poetry run pre-commit install --config $(PRE_COMMIT_CONFIG_PATH)
@echo "$(GREEN)Pre-commit hooks installed successfully.$(RESET)"
lint-backend:
@echo "$(YELLOW)Running linters...$(RESET)"
@poetry run pre-commit run --files opendevin/**/* agenthub/**/* evaluation/**/* --show-diff-on-failure --config $(PRECOMMIT_CONFIG_PATH)
@poetry run pre-commit run --files opendevin/**/* agenthub/**/* evaluation/**/* --show-diff-on-failure --config $(PRE_COMMIT_CONFIG_PATH)
lint-frontend:
@echo "$(YELLOW)Running linters for frontend...$(RESET)"
@@ -204,11 +208,10 @@ start-frontend:
@echo "$(YELLOW)Starting frontend...$(RESET)"
@cd frontend && VITE_BACKEND_HOST=$(BACKEND_HOST) VITE_FRONTEND_PORT=$(FRONTEND_PORT) npm run start
# Run the app
run:
@echo "$(YELLOW)Running the app...$(RESET)"
# Common setup for running the app (non-callable)
_run_setup:
@if [ "$(OS)" = "Windows_NT" ]; then \
echo "$(RED)`make run` is not supported on Windows. Please run `make start-frontend` and `make start-backend` separately.$(RESET)"; \
echo "$(RED) Windows is not supported, use WSL instead!$(RESET)"; \
exit 1; \
fi
@mkdir -p logs
@@ -217,9 +220,21 @@ run:
@echo "$(YELLOW)Waiting for the backend to start...$(RESET)"
@until nc -z localhost $(BACKEND_PORT); do sleep 0.1; done
@echo "$(GREEN)Backend started successfully.$(RESET)"
# Run the app (standard mode)
run:
@echo "$(YELLOW)Running the app...$(RESET)"
@$(MAKE) -s _run_setup
@cd frontend && echo "$(BLUE)Starting frontend with npm...$(RESET)" && npm run start -- --port $(FRONTEND_PORT)
@echo "$(GREEN)Application started successfully.$(RESET)"
# Run the app (WSL mode)
run-wsl:
@echo "$(YELLOW)Running the app in WSL mode...$(RESET)"
@$(MAKE) -s _run_setup
@cd frontend && echo "$(BLUE)Starting frontend with npm (WSL mode)...$(RESET)" && npm run dev_wsl -- --port $(FRONTEND_PORT)
@echo "$(GREEN)Application started successfully in WSL mode.$(RESET)"
# Setup config.toml
setup-config:
@echo "$(YELLOW)Setting up config.toml...$(RESET)"
@@ -234,8 +249,8 @@ setup-config-prompts:
workspace_dir=$${workspace_dir:-$(DEFAULT_WORKSPACE_DIR)}; \
echo "workspace_base=\"$$workspace_dir\"" >> $(CONFIG_FILE).tmp
@read -p "Do you want to persist the sandbox container? [true/false] [default: true]: " persist_sandbox; \
persist_sandbox=$${persist_sandbox:-true}; \
@read -p "Do you want to persist the sandbox container? [true/false] [default: false]: " persist_sandbox; \
persist_sandbox=$${persist_sandbox:-false}; \
if [ "$$persist_sandbox" = "true" ]; then \
read -p "Enter a password for the sandbox container: " ssh_password; \
echo "ssh_password=\"$$ssh_password\"" >> $(CONFIG_FILE).tmp; \
@@ -304,4 +319,4 @@ help:
@echo " $(GREEN)help$(RESET) - Display this help message, providing information on available targets."
# Phony targets
.PHONY: build check-dependencies check-python check-npm check-docker check-poetry pull-docker-image install-python-dependencies install-frontend-dependencies install-precommit-hooks lint start-backend start-frontend run setup-config setup-config-prompts help
.PHONY: build check-dependencies check-python check-npm check-docker check-poetry pull-docker-image install-python-dependencies install-frontend-dependencies install-pre-commit-hooks lint start-backend start-frontend run run-wsl setup-config setup-config-prompts help
+22 -17
View File
@@ -26,8 +26,6 @@
<br/>
<a href="https://join.slack.com/t/opendevin/shared_invite/zt-2i1iqdag6-bVmvamiPA9EZUu7oCO6KhA"><img src="https://img.shields.io/badge/Slack-Join%20Us-red?logo=slack&logoColor=white&style=for-the-badge" alt="Join our Slack community"></a>
<a href="https://discord.gg/ESHStjSjD4"><img src="https://img.shields.io/badge/Discord-Join%20Us-purple?logo=discord&logoColor=white&style=for-the-badge" alt="Join our Discord community"></a>
<br/>
<a href="https://huggingface.co/spaces/OpenDevin/evaluation"><img src="https://img.shields.io/badge/SWE--bench%20Lite-25.0%25-green?style=for-the-badge" alt="SWE-bench "></a>
<a href="https://codecov.io/github/opendevin/opendevin?branch=main"><img alt="CodeCov" src="https://img.shields.io/codecov/c/github/opendevin/opendevin?style=for-the-badge"></a>
</div>
@@ -35,7 +33,8 @@
<div align="center">
<img src="./docs/static/img/logo.png" alt="Logo" width="200" height="200">
<h1 align="center">OpenDevin: Code Less, Make More</h1>
<a href="https://opendevin.github.io/OpenDevin/"><img src="https://img.shields.io/badge/Documentation-OpenDevin-blue?logo=googledocs&logoColor=white&style=for-the-badge" alt="Check out the documentation"></a>
<a href="https://opendevin.github.io/OpenDevin/modules/usage/intro"><img src="https://img.shields.io/badge/Documentation-OpenDevin-blue?logo=googledocs&logoColor=white&style=for-the-badge" alt="Check out the documentation"></a>
<a href="https://huggingface.co/spaces/OpenDevin/evaluation"><img src="https://img.shields.io/badge/Evaluation-Benchmark%20on%20HF%20Space-green?style=for-the-badge" alt="Evaluation Benchmark"></a>
</div>
<hr>
@@ -55,29 +54,40 @@ To start OpenDevin in a docker container, run the following commands in your ter
> When you run the following command, files in `./workspace` may be modified or deleted.
```bash
OPENDEVIN_WORKSPACE=$(pwd)/workspace
WORKSPACE_BASE=$(pwd)/workspace
docker run -it \
--pull=always \
-e SANDBOX_USER_ID=$(id -u) \
-e PERSIST_SANDBOX="true" \
-e SSH_PASSWORD="make something up here" \
-e WORKSPACE_MOUNT_PATH=$OPENDEVIN_WORKSPACE \
-v $OPENDEVIN_WORKSPACE:/opt/workspace_base \
-e WORKSPACE_MOUNT_PATH=$WORKSPACE_BASE \
-v $WORKSPACE_BASE:/opt/workspace_base \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 3000:3000 \
--add-host host.docker.internal:host-gateway \
--name opendevin-app-$(date +%Y%m%d%H%M%S) \
ghcr.io/opendevin/opendevin:0.6
ghcr.io/opendevin/opendevin
```
You'll find OpenDevin running at [http://localhost:3000](http://localhost:3000) with access to `./workspace`. To have OpenDevin operate on your code, place it in `./workspace`.
> [!NOTE]
> By default, this command pulls the `latest` tag, which represents the most recent release of OpenDevin. You have other options as well:
> - For a specific release version, use `ghcr.io/opendevin/opendevin:<OpenDevin_version>` (replace <OpenDevin_version> with the desired version number).
> - For the most up-to-date development version, use `ghcr.io/opendevin/opendevin:main`. This version may be **(unstable!)** and is recommended for testing or development purposes only.
>
> Choose the tag that best suits your needs based on stability requirements and desired features.
You'll find OpenDevin running at [http://localhost:3000](http://localhost:3000) with access to `./workspace`. To have OpenDevin operate on your code, place it in `./workspace`.
OpenDevin will only have access to this workspace folder. The rest of your system will not be affected as it runs in a secured docker sandbox.
Upon opening OpenDevin, you must select the appropriate `Model` and enter the `API Key` within the settings that should pop up automatically. These can be set at any time by selecting
the `Settings` button (gear icon) in the UI. If the required `Model` does not exist in the list, you can manually enter it in the text box.
For the development workflow, see [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md).
Are you having trouble? Check out our [Troubleshooting Guide](https://opendevin.github.io/OpenDevin/modules/usage/troubleshooting).
## 🚀 Documentation
To learn more about the project, and for tips on using OpenDevin,
**check out our [documentation](https://opendevin.github.io/OpenDevin/)**.
**check out our [documentation](https://opendevin.github.io/OpenDevin/modules/usage/intro)**.
There you'll find resources on how to use different LLM providers (like ollama and Anthropic's Claude),
troubleshooting resources, and advanced configuration options.
@@ -103,11 +113,6 @@ Let's make software engineering better together!
- [Discord server](https://discord.gg/ESHStjSjD4) - This is a community-run server for general discussion, questions, and feedback.
## 📈 Progress
<p align="center">
<a href="https://www.swebench.com/lite.html">
<img src="/docs/static/img/results.png" alt="SWE-Bench Lite Score" width="500" height="auto">
</a>
</p>
<p align="center">
<a href="https://star-history.com/#OpenDevin/OpenDevin&Date">
@@ -141,4 +146,4 @@ Distributed under the MIT License. See [`LICENSE`](./LICENSE) for more informati
howpublished = {\url{https://github.com/OpenDevin/OpenDevin}},
note = {Accessed: ENTER THE DATE YOU ACCESSED THE PROJECT}
}
```
```
+1 -16
View File
@@ -29,12 +29,10 @@ The `state` contains:
Here is a list of available Actions, which can be returned by `agent.step()`:
- [`CmdRunAction`](../opendevin/events/action/commands.py) - Runs a command inside a sandboxed terminal
- [`CmdKillAction`](../opendevin/events/action/commands.py) - Kills a background command
- [`IPythonRunCellAction`](../opendevin/events/action/commands.py) - Execute a block of Python code interactively (in Jupyter notebook) and receives `CmdOutputObservation`. Requires setting up `jupyter` [plugin](../opendevin/runtime/plugins) as a requirement.
- [`FileReadAction`](../opendevin/events/action/files.py) - Reads the content of a file
- [`FileWriteAction`](../opendevin/events/action/files.py) - Writes new content to a file
- [`BrowseURLAction`](../opendevin/events/action/browse.py) - Gets the content of a URL
- [`AgentRecallAction`](../opendevin/events/action/agent.py) - Searches memory (e.g. a vector database)
- [`AddTaskAction`](../opendevin/events/action/tasks.py) - Adds a subtask to the plan
- [`ModifyTaskAction`](../opendevin/events/action/tasks.py) - Changes the state of a subtask.
- [`AgentFinishAction`](../opendevin/events/action/agent.py) - Stops the control loop, allowing the user/delegator agent to enter a new task
@@ -47,8 +45,7 @@ You can use `action.to_dict()` and `action_from_dict` to serialize and deseriali
## Observations
There are also several types of Observations. These are typically available in the step following the corresponding Action.
But they may also appear as a result of asynchronous events (e.g. a message from the user, logs from a command running
in the background).
But they may also appear as a result of asynchronous events (e.g. a message from the user).
Here is a list of available Observations:
@@ -56,7 +53,6 @@ Here is a list of available Observations:
- [`BrowserOutputObservation`](../opendevin/events/observation/browse.py)
- [`FileReadObservation`](../opendevin/events/observation/files.py)
- [`FileWriteObservation`](../opendevin/events/observation/files.py)
- [`AgentRecallObservation`](../opendevin/events/observation/recall.py)
- [`ErrorObservation`](../opendevin/events/observation/error.py)
- [`SuccessObservation`](../opendevin/events/observation/success.py)
@@ -74,14 +70,3 @@ def step(self, state: "State") -> "Action"
`step` moves the agent forward one step towards its goal. This probably means
sending a prompt to the LLM, then parsing the response into an `Action`.
### `search_memory`
```
def search_memory(self, query: str) -> list[str]:
```
`search_memory` should return a list of events that match the query. This will be used
for the `recall` action.
You can optionally just return `[]` for this method, meaning the agent has no long-term memory.
-405
View File
@@ -1,405 +0,0 @@
# GPT-4 Examples:
### Notes:
- The final result is not completed as I only gave it 10 steps, but shows very promising use of the tools.
- This whole run costs: $0.35 using gpt-4-turbo
- Input Tokens: 29,734
- Completion Tokens: 1,636
Interesting things:
1. Using comments as an initial placeholder for the next move and then having that output in memory is a very cool behavior.
2. It makes very small iterative changes due to the prompting.
3. A lot of the logic for the game is present but not put together well
4. There are a lot of odd quirks with the code that it has written.
5. Some odd agent behavior with putting random lines of code in random places?
6. The structure of the feedback loop needs some improvement so the agent can see the changes it has made better in context.
7. Might be beneficial to add in some line numbers so the model knows what lines it is looking at.
```
==============
STEP 0
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=91, prompt_tokens=2047, total_tokens=2138)
==== RAW OUTPUT ====
Thoughts:
To start implementing Conway's Game of Life in Python using the pygame library, I first need to set up a Python script file where I can write the code. I'll create a new Python file named `game_of_life.py`. This file will serve as the base for writing the game logic and using pygame functionalities.
Action:
"""
write game_of_life.py # This is a placeholder for the game of life implementation using pygame.
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='# This is a placeholder for the game of life implementation using pygame.', start=0, end=-1, thoughts="Thoughts:\nTo start implementing Conway's Game of Life in Python using the pygame library, I first need to set up a Python script file where I can write the code. I'll create a new Python file named `game_of_life.py`. This file will serve as the base for writing the game logic and using pygame functionalities.\n\nAction:", action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 0:-1.
==============
STEP 1
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=101, prompt_tokens=2351, total_tokens=2452)
==== RAW OUTPUT ====
Thoughts:
The placeholder for the game of life implementation using pygame has been set up in the file `game_of_life.py`. The next step is to start implementing the basic structure of the game. This includes importing necessary libraries, initializing the pygame module, setting up the display window, and defining the main game loop. I will begin by importing pygame and initializing it, which are the foundational steps for any pygame application.
Action:
"""
edit 0 0 import pygame
pygame.init()
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='import pygame\n\npygame.init()', start=0, end=0, thoughts='Thoughts:\nThe placeholder for the game of life implementation using pygame has been set up in the file `game_of_life.py`. The next step is to start implementing the basic structure of the game. This includes importing necessary libraries, initializing the pygame module, setting up the display window, and defining the main game loop. I will begin by importing pygame and initializing it, which are the foundational steps for any pygame application.\n\nAction:', action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 0:0.
==============
STEP 2
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=123, prompt_tokens=2552, total_tokens=2675)
==== RAW OUTPUT ====
Thoughts:
Previously, I initialized pygame in the `game_of_life.py` file. The next step is to set up the display window for the game. This involves defining the size of the window and setting the title. This is crucial as it provides the visual interface for the game. I will also add a comment to indicate that this is the setup for the display window.
Action:
"""
edit 2 2
# Setup the display window
screen_size = (800, 600)
screen = pygame.display.set_mode(screen_size)
pygame.display.set_caption("Conway's Game of Life")
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='# Setup the display window\nscreen_size = (800, 600)\nscreen = pygame.display.set_mode(screen_size)\npygame.display.set_caption("Conway\'s Game of Life")', start=2, end=2, thoughts='Thoughts:\nPreviously, I initialized pygame in the `game_of_life.py` file. The next step is to set up the display window for the game. This involves defining the size of the window and setting the title. This is crucial as it provides the visual interface for the game. I will also add a comment to indicate that this is the setup for the display window.\n\nAction:', action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 2:2.
==============
STEP 3
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=144, prompt_tokens=2777, total_tokens=2921)
==== RAW OUTPUT ====
Thoughts:
So far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame and setting up the display window. The next step is to define the main game loop where the game's logic will be continuously updated and rendered. This loop will handle events, update the game state, and redraw the screen. I will start by defining the main loop structure and a basic event handling mechanism to allow for quitting the game when the user presses the close button.
Action:
"""
edit 5 5 # Main game loop
running = True
while running:
for event in pygame.event.get():
if event.type == pygame.QUIT:
running = False
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='# Main game loop\nrunning = True\nwhile running:\n for event in pygame.event.get():\n if event.type == pygame.QUIT:\n running = False', start=5, end=5, thoughts="Thoughts:\nSo far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame and setting up the display window. The next step is to define the main game loop where the game's logic will be continuously updated and rendered. This loop will handle events, update the game state, and redraw the screen. I will start by defining the main loop structure and a basic event handling mechanism to allow for quitting the game when the user presses the close button.\n\nAction:", action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 5:5.
==============
STEP 4
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=222, prompt_tokens=3022, total_tokens=3244)
==== RAW OUTPUT ====
Thoughts:
So far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, and defining the main game loop. The next step is to implement the logic for Conway's Game of Life. This involves creating a grid to represent the game state, initializing it with a random pattern, and defining the rules for cell survival, birth, and death. I will start by defining the grid size and initializing the grid with random states.
Action:
"""
edit 7 7 # Define grid size and initialize grid
grid_size = (50, 50) # Grid size as a tuple (rows, columns)
grid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])] # Initialize grid with zeros
import random
for y in range(grid_size[0]):
for x in range(grid_size[1]):
grid[y][x] = random.randint(0, 1) # Randomly assign cells as alive (1) or dead (0)
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='# Define grid size and initialize grid\ngrid_size = (50, 50) # Grid size as a tuple (rows, columns)\ngrid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])] # Initialize grid with zeros\nimport random\nfor y in range(grid_size[0]):\n for x in range(grid_size[1]):\n grid[y][x] = random.randint(0, 1) # Randomly assign cells as alive (1) or dead (0)', start=7, end=7, thoughts="Thoughts:\nSo far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, and defining the main game loop. The next step is to implement the logic for Conway's Game of Life. This involves creating a grid to represent the game state, initializing it with a random pattern, and defining the rules for cell survival, birth, and death. I will start by defining the grid size and initializing the grid with random states.\n\nAction:", action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 7:7.
==============
STEP 5
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=234, prompt_tokens=3158, total_tokens=3392)
==== RAW OUTPUT ====
Thoughts:
So far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, and initializing the grid with random states. The next step is to implement the rules for Conway's Game of Life. This involves checking each cell's neighbors to determine its next state based on the number of alive neighbors. I will start by defining a function to count the alive neighbors for a given cell.
Action:
"""
edit 9 9 def count_alive_neighbors(grid, x, y):
directions = [(-1, -1), (-1, 0), (-1, 1), (0, -1), (0, 1), (1, -1), (1, 0), (1, 1)]
count = 0
for dx, dy in directions:
nx, ny = x + dx, y + dy
if 0 <= nx < grid_size[1] and 0 <= ny < grid_size[0]:
count += grid[ny][nx]
return count
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='def count_alive_neighbors(grid, x, y):\n directions = [(-1, -1), (-1, 0), (-1, 1), (0, -1), (0, 1), (1, -1), (1, 0), (1, 1)]\n count = 0\n for dx, dy in directions:\n nx, ny = x + dx, y + dy\n if 0 <= nx < grid_size[1] and 0 <= ny < grid_size[0]:\n count += grid[ny][nx]\n return count', start=9, end=9, thoughts="Thoughts:\nSo far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, and initializing the grid with random states. The next step is to implement the rules for Conway's Game of Life. This involves checking each cell's neighbors to determine its next state based on the number of alive neighbors. I will start by defining a function to count the alive neighbors for a given cell.\n\nAction:", action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 9:9.
==============
STEP 6
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=257, prompt_tokens=3290, total_tokens=3547)
==== RAW OUTPUT ====
Thoughts:
So far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating a function to count alive neighbors. The next step is to implement the rules for Conway's Game of Life. This involves updating the grid state based on the number of alive neighbors each cell has. I will define a function `update_grid` that applies the rules of the game to update the grid for the next generation.
Action:
"""
write game_of_life.py def update_grid(grid):
new_grid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])]
for y in range(grid_size[0]):
for x in range(grid_size[1]):
alive_neighbors = count_alive_neighbors(grid, x, y)
if grid[y][x] == 1 and (alive_neighbors == 2 or alive_neighbors == 3):
new_grid[y][x] = 1
elif grid[y][x] == 0 and alive_neighbors == 3:
new_grid[y][x] = 1
return new_grid
11
11
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='def update_grid(grid):\n new_grid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])]\n for y in range(grid_size[0]):\n for x in range(grid_size[1]):\n alive_neighbors = count_alive_neighbors(grid, x, y)\n if grid[y][x] == 1 and (alive_neighbors == 2 or alive_neighbors == 3):\n new_grid[y][x] = 1\n elif grid[y][x] == 0 and alive_neighbors == 3:\n new_grid[y][x] = 1\n return new_grid', start=11, end=11, thoughts="Thoughts:\nSo far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating a function to count alive neighbors. The next step is to implement the rules for Conway's Game of Life. This involves updating the grid state based on the number of alive neighbors each cell has. I will define a function `update_grid` that applies the rules of the game to update the grid for the next generation.\n\nAction:", action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 11:11.
==============
STEP 7
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=124, prompt_tokens=3416, total_tokens=3540)
==== RAW OUTPUT ====
Thoughts:
So far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating functions to count alive neighbors and update the grid. The next step is to integrate the `update_grid` function into the main game loop to continuously update the state of the grid based on the game rules. This will allow the game to evolve over time, showing different generations of the cell population.
Action:
"""
edit 6 6
grid = update_grid(grid)
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='grid = update_grid(grid)', start=6, end=6, thoughts="Thoughts:\nSo far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating functions to count alive neighbors and update the grid. The next step is to integrate the `update_grid` function into the main game loop to continuously update the state of the grid based on the game rules. This will allow the game to evolve over time, showing different generations of the cell population.\n\nAction:", action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 6:6.
==============
STEP 8
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=166, prompt_tokens=3393, total_tokens=3559)
==== RAW OUTPUT ====
Thoughts:
So far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating functions to count alive neighbors and update the grid. The next step is to integrate the `update_grid` function into the main game loop to continuously update the state of the grid based on the game rules. This will allow the game to evolve over time, showing different generations of the cell population. I need to ensure that the `update_grid` function is called within the main game loop and that the display is updated accordingly. I will start by checking the current content of the main game loop to see where to integrate the `update_grid` function.
Action:
"""
goto 6
"""
==== END RAW ====
ACTION:
FileReadAction(path='game_of_life.py', start=6, end=106, thoughts="Thoughts:\nSo far, I have set up the basic structure for Conway's Game of Life using pygame, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating functions to count alive neighbors and update the grid. The next step is to integrate the `update_grid` function into the main game loop to continuously update the state of the grid based on the game rules. This will allow the game to evolve over time, showing different generations of the cell population. I need to ensure that the `update_grid` function is called within the main game loop and that the display is updated accordingly. I will start by checking the current content of the main game loop to see where to integrate the `update_grid` function.\n\nAction:", action=<ActionType.READ: 'read'>)
OBSERVATION:
grid = update_grid(grid)
running = True
# Define grid size and initialize grid
grid_size = (50, 50) # Grid size as a tuple (rows, columns)
def count_alive_neighbors(grid, x, y):
directions = [(-1, -1), (-1, 0), (-1, 1), (0, -1), (0, 1), (1, -1), (1, 0), (1, 1)]
def update_grid(grid):
new_grid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])]
for y in range(grid_size[0]):
for x in range(grid_size[1]):
alive_neighbors = count_alive_neighbors(grid, x, y)
if grid[y][x] == 1 and (alive_neighbors == 2 or alive_neighbors == 3):
new_grid[y][x] = 1
elif grid[y][x] == 0 and alive_neighbors == 3:
new_grid[y][x] = 1
return new_grid
count = 0
for dx, dy in directions:
nx, ny = x + dx, y + dy
if 0 <= nx < grid_size[1] and 0 <= ny < grid_size[0]:
count += grid[ny][nx]
return count
grid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])] # Initialize grid with zeros
import random
for y in range(grid_size[0]):
for x in range(grid_size[1]):
grid[y][x] = random.randint(0, 1) # Randomly assign cells as alive (1) or dead (0)
while running:
for event in pygame.event.get():
if event.type == pygame.QUIT:
running = False
pygame.display.set_caption("Conway's Game of Life")
pygame.init()
==============
STEP 9
PLAN:
Please make conway's game of life in python using the pygame library. Be sure to use the tools in your workspace to help you out.
Usage(completion_tokens=174, prompt_tokens=3728, total_tokens=3902)
==== RAW OUTPUT ====
Thoughts:
From the previous actions, it seems that the basic structure for Conway's Game of Life using pygame has been set up, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating functions to count alive neighbors and update the grid. The next step is to ensure that the game loop properly updates the grid and refreshes the display to show the new state of the grid in each iteration. This involves calling `pygame.display.update()` to update the entire screen and using `pygame.time.Clock()` to control the frame rate. I will start by adding these necessary updates to the game loop.
Action:
"""
edit 95 95
pygame.display.update()
clock = pygame.time.Clock()
clock.tick(10) # Control the frame rate (10 frames per second)
"""
==== END RAW ====
ACTION:
FileWriteAction(path='game_of_life.py', content='pygame.display.update()\n clock = pygame.time.Clock()\n clock.tick(10) # Control the frame rate (10 frames per second)', start=95, end=95, thoughts="Thoughts:\nFrom the previous actions, it seems that the basic structure for Conway's Game of Life using pygame has been set up, including initializing pygame, setting up the display window, defining the main game loop, initializing the grid with random states, and creating functions to count alive neighbors and update the grid. The next step is to ensure that the game loop properly updates the grid and refreshes the display to show the new state of the grid in each iteration. This involves calling `pygame.display.update()` to update the entire screen and using `pygame.time.Clock()` to control the frame rate. I will start by adding these necessary updates to the game loop.\n\nAction:", action=<ActionType.WRITE: 'write'>)
OBSERVATION:
WRITE OPERATION:
You have written to "game_of_life.py" on these lines: 95:95.
16:43:55 - opendevin:INFO: agent_controller.py:125 - Exited before finishing the task.
```
## Result:
```
import pygame
# Setup the display window
screen_size = (800, 600)
screen = pygame.display.set_mode(screen_size)
# Main game loop
grid = update_grid(grid)
running = True
# Define grid size and initialize grid
grid_size = (50, 50) # Grid size as a tuple (rows, columns)
def count_alive_neighbors(grid, x, y):
directions = [(-1, -1), (-1, 0), (-1, 1), (0, -1), (0, 1), (1, -1), (1, 0), (1, 1)]
def update_grid(grid):
new_grid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])]
for y in range(grid_size[0]):
for x in range(grid_size[1]):
alive_neighbors = count_alive_neighbors(grid, x, y)
if grid[y][x] == 1 and (alive_neighbors == 2 or alive_neighbors == 3):
new_grid[y][x] = 1
elif grid[y][x] == 0 and alive_neighbors == 3:
new_grid[y][x] = 1
return new_grid
count = 0
for dx, dy in directions:
nx, ny = x + dx, y + dy
if 0 <= nx < grid_size[1] and 0 <= ny < grid_size[0]:
count += grid[ny][nx]
return count
grid = [[0 for x in range(grid_size[1])] for y in range(grid_size[0])] # Initialize grid with zeros
import random
for y in range(grid_size[0]):
for x in range(grid_size[1]):
grid[y][x] = random.randint(0, 1) # Randomly assign cells as alive (1) or dead (0)
while running:
for event in pygame.event.get():
if event.type == pygame.QUIT:
running = False
pygame.display.set_caption("Conway's Game of Life")
pygame.init()
# This is a placeholder for the game of life implementation using pygame.pygame.display.update()
clock = pygame.time.Clock()
clock.tick(10) # Control the frame rate (10 frames per second)
```
-5
View File
@@ -1,5 +0,0 @@
from opendevin.controller.agent import Agent
from .agent import SWEAgent
Agent.register('SWEAgent', SWEAgent)
-117
View File
@@ -1,117 +0,0 @@
from opendevin.controller.agent import Agent
from opendevin.controller.state.state import State
from opendevin.events.action import (
Action,
FileReadAction,
FileWriteAction,
MessageAction,
)
from opendevin.events.serialization.event import event_to_memory
from opendevin.llm.llm import LLM
from opendevin.runtime.tools import RuntimeTool
from .parser import parse_command
from .prompts import (
CONTEXT_PROMPT,
MEMORY_FORMAT,
NO_ACTION,
STEP_PROMPT,
SYSTEM_MESSAGE,
)
class SWEAgent(Agent):
VERSION = '1.0'
DEPRECATED = True
"""
An attempt to recreate swe_agent with output parsing, prompting style, and Application Computer Interface (ACI).
SWE-agent includes ACI functions like 'goto', 'search_for', 'edit', 'scroll', 'run'
"""
runtime_tools: list[RuntimeTool] = [RuntimeTool.BROWSER]
def __init__(self, llm: LLM):
super().__init__(llm)
self.memory_window = 4
self.max_retries = 2
self.cur_file: str = ''
self.cur_line: int = 0
def _think_act(self, messages: list[dict]) -> tuple[Action, str]:
resp = self.llm.do_completion(
messages=messages,
temperature=0.05,
)
action_resp = resp['choices'][0]['message']['content']
print(f"\033[1m\033[91m{resp['usage']}\033[0m")
print(
'\n==== RAW OUTPUT ====',
f'\033[96m{action_resp}\033[0m',
'==== END RAW ====\n',
sep='\n',
)
return parse_command(action_resp, self.cur_file, self.cur_line)
def _update(self, action: Action) -> None:
if isinstance(action, (FileReadAction, FileWriteAction)):
self.cur_file = action.path
self.cur_line = action.start
def step(self, state: State) -> Action:
"""
SWE-Agent step:
1. Get context - past actions, custom commands, current step
2. Perform think-act - prompt model for action and reasoning
3. Catch errors - ensure model takes action (5 attempts max)
"""
# retrieve short term memories from state.history, up to memory_window
memory_window = min(self.memory_window, len(state.history))
running_memory: list[str] = []
for prev_action, obs in state.history[-memory_window:]:
running_memory.append(
MEMORY_FORMAT(event_to_memory(prev_action), event_to_memory(obs))
)
goal = state.get_current_user_intent()
# always in the prompt if they exist: file and line
prompt = STEP_PROMPT(goal, self.cur_file, self.cur_line)
# prepare messages
msgs = [
{'content': SYSTEM_MESSAGE, 'role': 'system'},
{'content': prompt, 'role': 'user'},
]
# insert memories
if len(running_memory) > 0:
context = CONTEXT_PROMPT(running_memory, self.memory_window)
msgs.insert(1, {'content': context, 'role': 'user'})
# clrs = [''] * (len(msgs)-2) + ['\033[0;36m', '\033[0;35m']
# print('\n\n'.join([c+m['content']+'\033[0m' for c, m in zip(clrs, msgs)]))
# send it over
action, thought = self._think_act(messages=msgs)
# be robust with malformed responses
start_msg_len = len(msgs)
while not action and len(msgs) < self.max_retries + start_msg_len:
error = NO_ACTION(thought)
error_msg = {'content': error, 'role': 'user'}
msgs.append(error_msg)
action, thought = self._think_act(messages=msgs)
if not action:
action = MessageAction(thought)
self._update(action)
self.latest_action = action
return action
def search_memory(self, query: str) -> list[str]:
# return [item for item in self.running_memory if query in item]
raise NotImplementedError('Search_memory not implemented currently')
def reset(self) -> None:
"""Used to reset the agent"""
super().reset()
-167
View File
@@ -1,167 +0,0 @@
import re
from opendevin.events.action import (
Action,
AgentFinishAction,
BrowseURLAction,
CmdRunAction,
FileReadAction,
FileWriteAction,
MessageAction,
)
from .prompts import COMMAND_USAGE, CUSTOM_DOCS
# commands: exit, read, write, browse, kill, search_file, search_dir
no_open_file_error = MessageAction(
'You are not currently in a file. You can use the read command to open a file and then use goto to navigate through it.'
)
def invalid_error(cmd, docs):
return f"""ERROR:
Invalid command structure for
```
{cmd}
```
You may have caused this error by having multiple commands within your command block.
If so, try again by running only one of the commands:
Try again using this format:
{COMMAND_USAGE[docs]}
"""
def get_action_from_string(
command_string: str, path: str, line: int, thoughts: str = ''
) -> Action | None:
"""
Parses the command string to find which command the agent wants to run
Converts the command into a proper Action and returns
"""
vars = command_string.split(' ')
cmd = vars[0]
args = [] if len(vars) == 1 else vars[1:]
if 'exit' == cmd:
return AgentFinishAction()
elif 'think' == cmd:
return MessageAction(' '.join(args))
elif 'scroll_up' == cmd:
if not path:
return no_open_file_error
return FileReadAction(path, line + 100, line + 200, thoughts)
elif 'scroll_down' == cmd:
if not path:
return no_open_file_error
return FileReadAction(path, line - 100, line, thoughts)
elif 'goto' == cmd:
if not path:
return no_open_file_error
rex = r'^goto\s+(\d+)$'
valid = re.match(rex, command_string)
if valid:
start = int(valid.group(1))
end = start + 100
return FileReadAction(path, start, end, thoughts)
else:
return MessageAction(invalid_error(command_string, 'goto'))
elif 'edit' == cmd:
if not path:
return no_open_file_error
rex = r'^edit\s+(\d+)\s+(-?\d+)\s+(\S.*)$'
valid = re.match(rex, command_string, re.DOTALL)
if valid:
start = int(valid.group(1))
end = int(valid.group(2))
change = valid.group(3)
if '"' == change[-1] and '"' == change[0]:
change = change[1:-1]
return FileWriteAction(path, change, start, end, thoughts)
else:
return MessageAction(invalid_error(command_string, 'edit'))
elif 'read' == cmd:
rex = r'^read\s+(\S+)(?:\s+(\d+))?(?:\s+(-?\d+))?$'
valid = re.match(rex, command_string, re.DOTALL)
if valid:
file = valid.group(1)
start_str = valid.group(2)
end_str = valid.group(3)
start = 0 if not start_str else int(start_str)
end = -1 if not end_str else int(end_str)
return FileReadAction(file, start, end, thoughts)
else:
return MessageAction(invalid_error(command_string, 'read'))
elif 'write' == cmd:
rex = r'^write\s+(\S+)\s+(.*?)\s*(\d+)?\s*(-?\d+)?$'
valid = re.match(rex, command_string, re.DOTALL)
if valid:
file = valid.group(1)
content = valid.group(2)
start_str = valid.group(3)
end_str = valid.group(4)
start = 0 if not start_str else int(start_str)
end = -1 if not end_str else int(end_str)
if '"' == content[-1] and '"' == content[0]:
content = content[1:-1]
return FileWriteAction(file, content, start, end, thoughts)
else:
return MessageAction(invalid_error(command_string, 'write'))
elif 'browse' == cmd:
return BrowseURLAction(args[0].strip())
elif cmd in ['search_file', 'search_dir', 'find_file']:
rex = r'^(search_file|search_dir|find_file)\s+(\S+)(?:\s+(\S+))?$'
valid = re.match(rex, command_string, re.DOTALL)
if valid:
return CmdRunAction(command_string)
else:
return MessageAction(
f'Invalid command structure for\n ```\n{command_string}\n```.\nTry again using this format:\n{CUSTOM_DOCS}'
)
else:
# check bash command
obs = str(CmdRunAction(f'type {cmd}'))
if obs.split(':')[-1].strip() == 'not found':
# echo not found error for llm
return MessageAction(content=obs)
else:
# run valid command
return CmdRunAction(command_string)
def parse_command(input_str: str, path: str, line: int):
"""
Parses a given string and separates the command (enclosed in triple backticks) from any accompanying text.
Args:
input_str (str): The input string to be parsed.
Returns:
tuple: A tuple containing the command and the accompanying text (if any).
"""
input_str = input_str.strip()
if '```' in input_str:
parts = input_str.split('```')
command_str = parts[1].strip()
ind = 2 if len(parts) > 2 else 1
accompanying_text = ''.join(parts[:-ind]).strip()
action = get_action_from_string(command_str, path, line, accompanying_text)
if action:
return action, accompanying_text
return None, input_str # used for retry
-208
View File
@@ -1,208 +0,0 @@
DEFAULT_COMMANDS_DICT = {
'exit': 'Executed when task is complete',
'read <file_name> [<start_line>] [<end_line>]': "Shows a given file's contents starting from <start_line> up to <end_line>. Default: start_line = 0, end_line = -1. By default the whole file will be read.",
'write <file> <changes> [<start_line>] [<end_line>]': 'Modifies a <file> by replacing the current lines between <start_line> and <end_line> with <changes>. Default start_line = 0 and end_line = -1. Calling this with no line args will replace the whole file.',
'browse <url>': 'Returns the text version of any url, this can be useful to look up documentation or finding issues on github',
'scroll_up': 'Takes no arguments. This will scroll up and show you the 100 lines above your current lines',
'scroll_down': 'Takes no arguments. This will scroll down and show you the 100 lines below your current lines',
'edit <start_line> <end_line> <changes>': 'This will modify lines in the currently open file. use start_line and end_line to designate which lines to change and then write the multiline changes. Set end_line to -1 to denote the end of the file',
'goto <line_num>': 'This will take you directly to a line and show you the 100 lines below it.',
'<bash_command> <args>': 'You can use any bash command you need (cd, ls, rm, grep, dir, mv, wget, git, zip, etc.) with their arguments included',
'pip install <package>': 'You can use this to import python packages. Make sure you include the correct package name when using this command.',
'ls': 'Use the ls command to view all the files in your current directory, this is a good starting point.',
'NOT ALLOWED': 'You cannot use interactive commands like python or node',
}
COMMAND_USAGE = {
'exit': 'Usage:\n```\nexit\n```\nExecuted when task is complete',
'read': "Args:\n<file_name> [<start_line>] [<end_line>]\nUsage:\n```\nread file.py\n```\nor\n```\nread example.py <start_line> <end_line>\n```\nShows a given file's contents starting from <start_line> up to <end_line>. Default: start_line = 0, end_line = -1. by default the whole file will be read.",
'write': 'Args:\n<file> <changes> [<start_line>] [<end_line>]\nUsage:\n```\nwrite "def main():\n print("This is line one")" 0 2\n```\nModifies a <file> by replacing the current lines between <start_line> and <end_line> with <changes>. Default start_line = 0 and end_line = -1. Calling this with no line args will replace the whole file.',
'edit': 'Args:\n<start_line> <end_line> <changes>\nUsage:\n```\nedit 0 1 import pandas as pd\n```\nThis will modify the current file you are in with the changes you make between the line numbers you designate',
'goto': 'Args:\n<line_num>\nUsage:\n```\ngoto <line_num>\n```\nThis will show you the 100 lines below and including the line you specify within your current file.',
'scroll_up': 'Usage:\n```\nscroll_up\n```\nThis will return the 100 lines above where you are currently at',
'scroll_down': 'Usage:\n```\nscroll_down\n```\nThis will return the 100 line below where you are currently at',
'browse': 'Args:\n<url>\nUsage:\n```\nbrowse https://github.com/OpenDevin/OpenDevin\n```\nThis will fetch the Text elements from the given url and show them to you.',
}
DEFAULT_COMMANDS = '\n'.join([k + ' - ' + v for k, v in DEFAULT_COMMANDS_DICT.items()])
# from opendevin.parse_commands import parse_command_file
# USE parse_command_file(filepath) to get the custom commands
CUSTOM_DOCS = None
CUSTOM_COMMANDS = f"""Custom bash commands:
{CUSTOM_DOCS}
"""
DOCUMENTATION = f"""DOCUMENTATION:
It is recommend that you use the commands provided for interacting with files and your directory because they have been specially built for you.
They will make it much easier for you to look at files and make changes. Using these commands will help you be better at your task.
You can open an file by using either the read or write operations.
- If a file already exists you should read it before making any changes. Use the `edit` command to make changes once you have read it.
- If you are creating a new file use the write command. Use the `edit` command to make changes once you have created the new file.
Commands:
{DEFAULT_COMMANDS}
{CUSTOM_COMMANDS}
The following commands require an open file to be used: edit, scroll_up, scroll_down, goto
To modify the current file use 'edit'. To move through the current file use 'goto' or 'scroll_up'/'scroll_down'
when using write and edit do not surround the code with any "" just write the code.
"""
GENERAL_GUIDELINES = """INSTRUCTIONS:
Now, you're going to solve this issue on your own. You can use any bash commands or custom commands you wish to complete your task. Edit all the files you need to and run any checks or tests that you want.
Remember, YOU CAN ONLY ENTER ONE COMMAND AT A TIME. You should always wait for feedback after every command.
When you're satisfied with all of the changes you've made, you can indicate that you are done by running the exit command.
Note however that you cannot use any interactive session commands (e.g. python, vim, node) in this environment, but you can write scripts and run them. E.g. you can write a python script and then run it with `python <script_name>.py`.
NOTE ABOUT THE write COMMAND: Indentation really matters! When editing a file, make sure to insert appropriate indentation before each line!
IMPORTANT TIPS:
1. Reproduce the bug: Always start by trying to replicate the bug that the issue discusses. If the issue includes code for reproducing the bug, we recommend that you re-implement that in your environment and run it to ensure you can reproduce the bug. Then, start trying to fix it. When you think you've fixed the bug, re-run the bug reproduction script to make sure that the issue has indeed been resolved.
If the bug reproduction script does not print anything when it successfully runs, we recommend adding a print("Script completed successfully, no errors.") command at the end of the file, so that you can be sure the script ran fine all the way through.
2. Try different commands: If you run a command and it doesn't work, try running a different command. A command that did not work once will not work the second time unless you modify it.
3. Navigate large files: If you open a file and need to get to an area around a specific line that is not in the first 100 lines, say line 583, you would use the 'read' command like this: 'read <file> 583'. This is a much faster way to read through the file.
4. Handle input files: If the bug reproduction script requires inputting/reading a specific file, such as 'buggy-input.png', and you'd like to understand how to input that file, conduct a search in the existing repository code to see whether someone else has already done that. Do this by running the command: 'search_dir "buggy-input.png"'. If that doesn't work, use the Linux 'find' command.
5. Understand your context: Always make sure to look at the currently open file and the current working directory. The currently open file might be in a different directory than the working directory.
6. Verify your edits: When editing files, it is easy to accidentally specify a wrong line number or to write code with incorrect indentation. Always check the code after you issue an edit to make sure that it reflects what you wanted to accomplish. If it didn't, issue another command to fix it.
7. Thoroughly test your solution: After making any changes to fix a bug, be sure to thoroughly test your solution to ensure the bug has been resolved. Re-run the bug reproduction script and verify that the issue has been addressed.
"""
RESPONSE_FORMAT = """RESPONSE FORMAT:
This is the format of the response you will make in order to solve the current issue.
You will be given multiple iterations to complete this task so break it into steps and solve them one by one.
Your output must contain the following:
- First, thoughts about what your next action should be and plan it out.
- You will have a memory of your thoughts so you can use this to remember things for the next step.
- Use your thoughts to think about what you are currently doing, what you have done on prior steps and how that relates to solving the problem.
- Second, create a piece of code that will execute your next action based on the thoughts you have.
- Remember that you can only have one action for each thought, do not include multiple actions.
Your code MUST be surrounded in triple back ticks EXACTLY like this:
```
<code>
```
Notes:
- Adhere to the format so that the program loop continues smoothly, it is very important to only give one command per output.
- DO NOT give more than one command within the triple backticks. This will just throw an error and nothing will happen as a result.
- Do not give multiple code blocks, if you do only the second one will be captured and run, this might give an error if the first one was necessary.
- To execute multiple commands you should write them down in your thoughts section so you can remember it on the next step and execute them then.
- The only commands you are not capable of executing are interactive commands like `python` or `node` by themselves.
- If you think that you have completed the task that has been given to you based on your previous actions and outputs then use ``` exit ``` as the command to let the system know that you are done.
- DO NOT make any copies of your previous memories, those will be provided to you at each step, making copies just wastes time and energy. Think smarter not harder.
- The write and edit commands requires proper indentation in the content section ex. `write hw.py def hello():\n print(\'Hello World\')` this is how you would have to format your write command.
- The white spaces matter as the code changes will be added to the code so they must have proper syntax.
This is a template using the format described above
Items in <> are suggestions for you, fill them out based on the context of the problem you are solving.
[ FORMAT ]
Thoughts:
<Provide clear and concise thoughts on the next step to take, highlighting any important details or context that should be remembered.>
<You can use multiple lines to express your thoughts>
Action:
```
<command> <params>
```
[ END FORMAT ]
Do not provide anything extra just your thought and action.
"""
SYSTEM_MESSAGE = f"""SYSTEM INFO:
You are an autonomous coding agent, here to provide solutions for coding issues.
You have been designed to assist with a wide range of programming tasks, from code editing and debugging to testing and deployment.
You have access to a variety of tools and commands that you can use to help you solve problems efficiently.
{GENERAL_GUIDELINES}
{DOCUMENTATION}
""".strip()
def NO_ACTION(latest):
return f"""
You did not include any action to take in your most recent output:
===== Output ======
{latest}
==== End Output ===
Remember these are the custom commands you can use:
{DOCUMENTATION}
Lets try that again, it is very important that you adhere to the output format
This time, be sure to use the exact format below, replacing anything in <> with the appropriate value(s):
{RESPONSE_FORMAT}
It is crucial you use the format provided as the output will be parsed automatically.
"""
def file_info(file: str, line: int):
if file:
return f"""CURRENT WORKSPACE:
Open File: {file} on line {line}
You can use these commands with the current file:
Navigation: `scroll_up`, `scroll_down`, and `goto <line>`
Modification: `edit <start_line> <end_line> <changes>`
"""
def STEP_PROMPT(task, file, line_num):
return f"""
{RESPONSE_FORMAT}
You are currently trying to complete this task:
{task}
{file_info(file, line_num)}
Keep all of the guidelines above in mind when you are thinking and making code.
Please come up with a thought and action based on your current task and latest steps.
Make sure that you do not repeat the same actions, there will not be any changes in result if you do not changes anything.
Be very strict about the formatting that you use and make sure you follow the guidelines.
NEVER output multiple commands. ONLY take ONE STEP at a time.
When you have completed your task run the "exit" command.
Begin with your thought about the next step and then come up with an action to perform your thought.
""".strip()
def unpack_dict(data: dict, restrict: list[str] | None = None):
lines = []
restrict = [] if restrict is None else restrict
for key, value in data.items():
if key in restrict:
continue
elif isinstance(value, dict):
nested_str = unpack_dict(value, restrict).replace('\n', '\n ')
val = f'{key}:' + '\n ' + f'{nested_str}'
lines.append(val)
else:
lines.append(f'{key}: {value}')
return '\n'.join(lines)
def MEMORY_FORMAT(act, obs):
return f"""
Previous Action:
{unpack_dict(act, ["content"])}
Output from Action:
{unpack_dict(obs)}
""".strip()
def CONTEXT_PROMPT(memory, window):
res = f'These are your past {window} actions:\n'
window_size = window if len(memory) > window else len(memory)
cur_mems = memory[-window_size:]
res += '===== Previous Actions =====\n'
for idx, mem in enumerate(cur_mems):
res += f'\nMemory {idx}:\n{mem}\n'
res += '======= End Actions =======\n'
res += 'Use these memories to provide additional context to the problem you are solving.\nRemember that you have already completed these steps so you do not need to perform them again.'
return res
-2
View File
@@ -9,7 +9,6 @@ load_dotenv()
from . import ( # noqa: E402
SWE_agent,
browsing_agent,
codeact_agent,
codeact_swe_agent,
@@ -24,7 +23,6 @@ __all__ = [
'codeact_agent',
'codeact_swe_agent',
'planner_agent',
'SWE_agent',
'delegator_agent',
'dummy_agent',
'browsing_agent',
+76 -108
View File
@@ -1,9 +1,9 @@
import ast
import os
from browsergym.core.action.highlevel import HighLevelActionSet
from browsergym.utils.obs import flatten_axtree_to_str
from agenthub.browsing_agent.response_parser import BrowsingResponseParser
from opendevin.controller.agent import Agent
from opendevin.controller.state.state import State
from opendevin.core.logger import opendevin_logger as logger
@@ -15,6 +15,7 @@ from opendevin.events.action import (
)
from opendevin.events.event import EventSource
from opendevin.events.observation import BrowserOutputObservation
from opendevin.events.observation.observation import Observation
from opendevin.llm.llm import LLM
from opendevin.runtime.plugins import (
PluginRequirement,
@@ -34,6 +35,56 @@ else:
EVAL_MODE = False
def get_error_prefix(last_browser_action: str) -> str:
return f'IMPORTANT! Last action is incorrect:\n{last_browser_action}\nThink again with the current observation of the page.\n'
def get_system_message(goal: str, action_space: str) -> str:
return f"""\
# Instructions
Review the current state of the page and all other information to find the best
possible next action to accomplish your goal. Your answer will be interpreted
and executed by a program, make sure to follow the formatting instructions.
# Goal:
{goal}
# Action Space
{action_space}
"""
CONCISE_INSTRUCTION = """\
Here is another example with chain of thought of a valid action when providing a concise answer to user:
"
In order to accomplish my goal I need to send the information asked back to the user. This page list the information of HP Inkjet Fax Machine, which is the product identified in the objective. Its price is $279.49. I will send a message back to user with the answer.
```send_msg_to_user("$279.49")```
"
"""
def get_prompt(error_prefix: str, cur_axtree_txt: str, prev_action_str: str) -> str:
prompt = f"""\
{error_prefix}
# Current Accessibility Tree:
{cur_axtree_txt}
# Previous Actions
{prev_action_str}
Here is an example with chain of thought of a valid action when clicking on a button:
"
In order to accomplish my goal I need to click on the button with bid 12
```click("12")```
"
""".strip()
if USE_CONCISE_ANSWER:
prompt += CONCISE_INSTRUCTION
return prompt
class BrowsingAgent(Agent):
VERSION = '1.0'
"""
@@ -42,6 +93,7 @@ class BrowsingAgent(Agent):
sandbox_plugins: list[PluginRequirement] = []
runtime_tools: list[RuntimeTool] = [RuntimeTool.BROWSER]
response_parser = BrowsingResponseParser()
def __init__(
self,
@@ -75,31 +127,6 @@ class BrowsingAgent(Agent):
self.cost_accumulator = 0
self.error_accumulator = 0
def parse_response(self, response: str) -> Action:
if '```' not in response:
# unexpected response format, message back to user
action_str = f'send_msg_to_user("""{response}""")'
return BrowseInteractiveAction(
browser_actions=action_str,
thought=response,
browsergym_send_msg_to_user=response,
)
thought = response.split('```')[0].strip()
action_str = response.split('```')[1].strip()
# handle send message to user function call in BrowserGym
msg_content = ''
for sub_action in action_str.split('\n'):
if 'send_msg_to_user(' in sub_action:
tree = ast.parse(sub_action)
args = tree.body[0].value.args # type: ignore
msg_content = args[0].value
return BrowseInteractiveAction(
browser_actions=action_str,
thought=thought,
browsergym_send_msg_to_user=msg_content,
)
def step(self, state: State) -> Action:
"""
Performs one step using the Browsing Agent.
@@ -113,9 +140,6 @@ class BrowsingAgent(Agent):
- MessageAction(content) - Message action to run (e.g. ask for clarification)
- AgentFinishAction() - end the interaction
"""
goal = state.get_current_user_intent()
if goal is None:
goal = state.inputs['task']
messages = []
prev_actions = []
cur_axtree_txt = ''
@@ -123,23 +147,21 @@ class BrowsingAgent(Agent):
last_obs = None
last_action = None
if EVAL_MODE and len(state.history) == 1:
if EVAL_MODE and len(state.history.get_events_as_list()) == 1:
# for webarena and miniwob++ eval, we need to retrieve the initial observation already in browser env
# initialize and retrieve the first observation by issuing an noop OP
# For non-benchmark browsing, the browser env starts with a blank page, and the agent is expected to first navigate to desired websites
return BrowseInteractiveAction(browser_actions='noop()')
for prev_action, obs in state.history:
if isinstance(prev_action, BrowseInteractiveAction):
prev_actions.append(prev_action.browser_actions)
last_obs = obs
last_action = prev_action
elif (
isinstance(prev_action, MessageAction)
and prev_action.source == EventSource.AGENT
):
# agent has responded, task finish.
return AgentFinishAction(outputs={'content': prev_action.content})
for event in state.history.get_events():
if isinstance(event, BrowseInteractiveAction):
prev_actions.append(event.browser_actions)
last_action = event
elif isinstance(event, MessageAction) and event.source == EventSource.AGENT:
# agent has responded, task finished.
return AgentFinishAction(outputs={'content': event.content})
elif isinstance(event, Observation):
last_obs = event
if EVAL_MODE:
prev_actions = prev_actions[1:] # remove the first noop action
@@ -156,7 +178,10 @@ class BrowsingAgent(Agent):
if isinstance(last_obs, BrowserOutputObservation):
if last_obs.error:
# add error recovery prompt prefix
error_prefix = f'IMPORTANT! Last action is incorrect:\n{last_obs.last_browser_action}\nThink again with the current observation of the page.\n'
error_prefix = get_error_prefix(last_obs.last_browser_action)
self.error_accumulator += 1
if self.error_accumulator > 5:
return MessageAction('Too many errors encountered. Task failed.')
try:
cur_axtree_txt = flatten_axtree_to_str(
last_obs.axtree_object,
@@ -170,78 +195,21 @@ class BrowsingAgent(Agent):
)
return MessageAction('Error encountered when browsing.')
if error_prefix:
self.error_accumulator += 1
if self.error_accumulator > 5:
return MessageAction('Too many errors encountered. Task failed.')
system_msg = f"""\
# Instructions
Review the current state of the page and all other information to find the best
possible next action to accomplish your goal. Your answer will be interpreted
and executed by a program, make sure to follow the formatting instructions.
# Goal:
{goal}
# Action Space
{self.action_space.describe(with_long_description=False, with_examples=True)}
"""
if (goal := state.get_current_user_intent()) is None:
goal = state.inputs['task']
system_msg = get_system_message(
goal,
self.action_space.describe(with_long_description=False, with_examples=True),
)
messages.append({'role': 'system', 'content': system_msg})
prompt = f"""\
{error_prefix}
# Current Accessibility Tree:
{cur_axtree_txt}
# Previous Actions
{prev_action_str}
Here is an example with chain of thought of a valid action when clicking on a button:
"
In order to accomplish my goal I need to click on the button with bid 12
```click("12")```
"
""".strip()
if USE_CONCISE_ANSWER:
concise_instruction = """\
Here is another example with chain of thought of a valid action when providing a concise answer to user:
"
In order to accomplish my goal I need to send the information asked back to the user. This page list the information of HP Inkjet Fax Machine, which is the product identified in the objective. Its price is $279.49. I will send a message back to user with the answer.
```send_msg_to_user("$279.49")```
"
"""
prompt += concise_instruction
prompt = get_prompt(error_prefix, cur_axtree_txt, prev_action_str)
messages.append({'role': 'user', 'content': prompt})
logger.debug(prompt)
response = self.llm.completion(
messages=messages,
temperature=0.0,
stop=[')```', ')\n```'],
)
self.log_cost(response)
action_resp = response['choices'][0]['message']['content'].strip()
if not action_resp.endswith('```'):
action_resp = action_resp + ')```'
logger.info(prompt)
logger.info(action_resp)
return self.parse_response(action_resp)
def search_memory(self, query: str) -> list[str]:
raise NotImplementedError('Implement this abstract method')
def log_cost(self, response):
# TODO: refactor to unified cost tracking
try:
cur_cost = self.llm.completion_cost(response)
except Exception:
cur_cost = 0
self.cost_accumulator += cur_cost
logger.info(
'Cost: %.2f USD | Accumulated Cost: %.2f USD',
cur_cost,
self.cost_accumulator,
)
return self.response_parser.parse(response)
@@ -0,0 +1,90 @@
import ast
from opendevin.controller.action_parser import ActionParser, ResponseParser
from opendevin.core.logger import opendevin_logger as logger
from opendevin.events.action import (
Action,
BrowseInteractiveAction,
)
class BrowsingResponseParser(ResponseParser):
def __init__(self):
# Need to pay attention to the item order in self.action_parsers
super().__init__()
self.action_parsers = [BrowsingActionParserMessage()]
self.default_parser = BrowsingActionParserBrowseInteractive()
def parse(self, response: str) -> Action:
action_str = self.parse_response(response)
return self.parse_action(action_str)
def parse_response(self, response) -> str:
action_str = response['choices'][0]['message']['content']
if action_str is None:
return ''
action_str = action_str.strip()
if not action_str.endswith('```'):
action_str = action_str + ')```'
logger.info(action_str)
return action_str
def parse_action(self, action_str: str) -> Action:
for action_parser in self.action_parsers:
if action_parser.check_condition(action_str):
return action_parser.parse(action_str)
return self.default_parser.parse(action_str)
class BrowsingActionParserMessage(ActionParser):
"""
Parser action:
- BrowseInteractiveAction(browser_actions) - unexpected response format, message back to user
"""
def __init__(
self,
):
pass
def check_condition(self, action_str: str) -> bool:
return '```' not in action_str
def parse(self, action_str: str) -> Action:
msg = f'send_msg_to_user("""{action_str}""")'
return BrowseInteractiveAction(
browser_actions=msg,
thought=action_str,
browsergym_send_msg_to_user=action_str,
)
class BrowsingActionParserBrowseInteractive(ActionParser):
"""
Parser action:
- BrowseInteractiveAction(browser_actions) - handle send message to user function call in BrowserGym
"""
def __init__(
self,
):
pass
def check_condition(self, action_str: str) -> bool:
return True
def parse(self, action_str: str) -> Action:
thought = action_str.split('```')[0].strip()
action_str = action_str.split('```')[1].strip()
msg_content = ''
for sub_action in action_str.split('\n'):
if 'send_msg_to_user(' in sub_action:
tree = ast.parse(sub_action)
args = tree.body[0].value.args # type: ignore
msg_content = args[0].value
return BrowseInteractiveAction(
browser_actions=action_str,
thought=thought,
browsergym_send_msg_to_user=msg_content,
)
+5 -4
View File
@@ -21,10 +21,9 @@ class CodeActResponseParser(ResponseParser):
- AgentFinishAction() - end the interaction
"""
def __init__(
self,
):
def __init__(self):
# Need pay attention to the item order in self.action_parsers
super().__init__()
self.action_parsers = [
CodeActActionParserFinish(),
CodeActActionParserCmdRun(),
@@ -33,12 +32,14 @@ class CodeActResponseParser(ResponseParser):
]
self.default_parser = CodeActActionParserMessage()
def parse(self, response: str) -> Action:
def parse(self, response) -> Action:
action_str = self.parse_response(response)
return self.parse_action(action_str)
def parse_response(self, response) -> str:
action = response.choices[0].message.content
if action is None:
return ''
for lang in ['bash', 'ipython', 'browse']:
if f'<execute_{lang}>' in action and f'</execute_{lang}>' not in action:
action += f'</execute_{lang}>'
+57 -55
View File
@@ -8,20 +8,21 @@ from agenthub.codeact_agent.prompt import (
)
from opendevin.controller.agent import Agent
from opendevin.controller.state.state import State
from opendevin.core.config import config
from opendevin.events.action import (
Action,
AgentDelegateAction,
AgentFinishAction,
BrowseInteractiveAction,
CmdRunAction,
IPythonRunCellAction,
MessageAction,
)
from opendevin.events.observation import (
AgentDelegateObservation,
BrowserOutputObservation,
CmdOutputObservation,
IPythonRunCellObservation,
)
from opendevin.events.serialization.event import truncate_content
from opendevin.llm.llm import LLM
from opendevin.runtime.plugins import (
AgentSkillsRequirement,
@@ -38,8 +39,8 @@ def action_to_str(action: Action) -> str:
return f'{action.thought}\n<execute_bash>\n{action.command}\n</execute_bash>'
elif isinstance(action, IPythonRunCellAction):
return f'{action.thought}\n<execute_ipython>\n{action.code}\n</execute_ipython>'
elif isinstance(action, BrowseInteractiveAction):
return f'{action.thought}\n<execute_browse>\n{action.browser_actions}\n</execute_browse>'
elif isinstance(action, AgentDelegateAction):
return f'{action.thought}\n<execute_browse>\n{action.inputs["task"]}\n</execute_browse>'
elif isinstance(action, MessageAction):
return action.content
return ''
@@ -47,7 +48,7 @@ def action_to_str(action: Action) -> str:
def get_action_message(action: Action) -> dict[str, str] | None:
if (
isinstance(action, BrowseInteractiveAction)
isinstance(action, AgentDelegateAction)
or isinstance(action, CmdRunAction)
or isinstance(action, IPythonRunCellAction)
or isinstance(action, MessageAction)
@@ -60,10 +61,13 @@ def get_action_message(action: Action) -> dict[str, str] | None:
def get_observation_message(obs) -> dict[str, str] | None:
max_message_chars = config.get_llm_config_from_agent(
'CodeActAgent'
).max_message_chars
if isinstance(obs, CmdOutputObservation):
content = 'OBSERVATION:\n' + truncate_observation(obs.content)
content = 'OBSERVATION:\n' + truncate_content(obs.content, max_message_chars)
content += (
f'\n[Command {obs.command_id} finished with exit code {obs.exit_code}]]'
f'\n[Command {obs.command_id} finished with exit code {obs.exit_code}]'
)
return {'role': 'user', 'content': content}
elif isinstance(obs, IPythonRunCellObservation):
@@ -76,31 +80,16 @@ def get_observation_message(obs) -> dict[str, str] | None:
'![image](data:image/png;base64, ...) already displayed to user'
)
content = '\n'.join(splitted)
content = truncate_observation(content)
return {'role': 'user', 'content': content}
elif isinstance(obs, BrowserOutputObservation):
content = 'OBSERVATION:\n' + truncate_observation(obs.content)
content = truncate_content(content, max_message_chars)
return {'role': 'user', 'content': content}
elif isinstance(obs, AgentDelegateObservation):
content = 'OBSERVATION:\n' + truncate_observation(str(obs.outputs))
content = 'OBSERVATION:\n' + truncate_content(
str(obs.outputs), max_message_chars
)
return {'role': 'user', 'content': content}
return None
def truncate_observation(observation: str, max_chars: int = 10_000) -> str:
"""
Truncate the middle of the observation if it is too long.
"""
if len(observation) <= max_chars:
return observation
half = max_chars // 2
return (
observation[:half]
+ '\n[... Observation truncated due to length ...]\n'
+ observation[-half:]
)
# FIXME: We can tweak these two settings to create MicroAgents specialized toward different area
def get_system_message() -> str:
if ENABLE_GITHUB:
@@ -114,7 +103,7 @@ def get_in_context_example() -> str:
class CodeActAgent(Agent):
VERSION = '1.6'
VERSION = '1.8'
"""
The Code Act Agent is a minimalist agent.
The agent works by passing the model a list of action-observation pairs and prompting the model to take the next step.
@@ -153,8 +142,8 @@ class CodeActAgent(Agent):
sandbox_plugins: list[PluginRequirement] = [
# NOTE: AgentSkillsRequirement need to go before JupyterRequirement, since
# AgentSkillsRequirement provides a lot of Python functions
# and it need to be initialized before Jupyter for Jupyter to use those functions.
# AgentSkillsRequirement provides a lot of Python functions,
# and it needs to be initialized before Jupyter for Jupyter to use those functions.
AgentSkillsRequirement(),
JupyterRequirement(),
]
@@ -190,7 +179,7 @@ class CodeActAgent(Agent):
This includes gathering info on previous steps and prompting the model to make a command to execute.
Parameters:
- state (State): used to get updated info and background commands
- state (State): used to get updated info
Returns:
- CmdRunAction(command) - bash command to run
@@ -199,29 +188,16 @@ class CodeActAgent(Agent):
- MessageAction(content) - Message action to run (e.g. ask for clarification)
- AgentFinishAction() - end the interaction
"""
messages: list[dict[str, str]] = [
{'role': 'system', 'content': self.system_message},
{'role': 'user', 'content': self.in_context_example},
]
for prev_action, obs in state.history:
action_message = get_action_message(prev_action)
if action_message:
messages.append(action_message)
# if we're done, go back
latest_user_message = state.history.get_last_user_message()
if latest_user_message and latest_user_message.strip() == '/exit':
return AgentFinishAction()
obs_message = get_observation_message(obs)
if obs_message:
messages.append(obs_message)
# prepare what we want to send to the LLM
messages: list[dict[str, str]] = self._get_messages(state)
latest_user_message = [m for m in messages if m['role'] == 'user'][-1]
if latest_user_message:
if latest_user_message['content'].strip() == '/exit':
return AgentFinishAction()
latest_user_message['content'] += (
f'\n\nENVIRONMENT REMINDER: You have {state.max_iterations - state.iteration} turns left to complete the task.'
)
response = self.llm.do_completion(
response = self.llm.completion(
messages=messages,
stop=[
'</execute_ipython>',
@@ -230,10 +206,36 @@ class CodeActAgent(Agent):
],
temperature=0.0,
)
state.num_of_chars += sum(
len(message['content']) for message in messages
) + len(response.choices[0].message.content)
return self.action_parser.parse(response)
def search_memory(self, query: str) -> list[str]:
raise NotImplementedError('Implement this abstract method')
def _get_messages(self, state: State) -> list[dict[str, str]]:
messages = [
{'role': 'system', 'content': self.system_message},
{'role': 'user', 'content': self.in_context_example},
]
for event in state.history.get_events():
# create a regular message from an event
message = (
get_action_message(event)
if isinstance(event, Action)
else get_observation_message(event)
)
# add regular message
if message:
messages.append(message)
# the latest user message is important:
# we want to remind the agent of the environment constraints
latest_user_message = next(
(m for m in reversed(messages) if m['role'] == 'user'), None
)
# add a reminder to the prompt
if latest_user_message:
latest_user_message['content'] += (
f'\n\nENVIRONMENT REMINDER: You have {state.max_iterations - state.iteration} turns left to complete the task. When finished reply with <finish></finish>'
)
return messages
+27 -7
View File
@@ -5,7 +5,7 @@ _AGENT_SKILLS_DOCS = AgentSkillsRequirement.documentation
COMMAND_DOCS = (
'\nApart from the standard Python library, the assistant can also use the following functions (already imported) in <execute_ipython> environment:\n'
f'{_AGENT_SKILLS_DOCS}'
"Please note that THE `edit_file` and `append_file` FUNCTIONS REQUIRE PROPER INDENTATION. If the assistant would like to add the line ' print(x)', it must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run."
"Please note that THE `edit_file_by_replace`, `append_file` and `insert_content_at_line` FUNCTIONS REQUIRE PROPER INDENTATION. If the assistant would like to add the line ' print(x)', it must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run."
)
# ======= SYSTEM MESSAGE =======
@@ -37,9 +37,11 @@ For example, to push a branch `my_branch` to the GitHub repo `owner/repo`:
If $GITHUB_TOKEN is not set, ask the user to set it."""
SYSTEM_SUFFIX = """Responses should be concise.
The assistant should attempt fewer things at a time instead of putting too much commands OR code in one "execute" block.
Include ONLY ONE <execute_ipython>, <execute_bash>, or <execute_browse> per response, unless the assistant is finished with the task or need more input or action from the user in order to proceed.
The assistant should attempt fewer things at a time instead of putting too many commands OR too much code in one "execute" block.
Include ONLY ONE <execute_ipython>, <execute_bash>, or <execute_browse> per response, unless the assistant is finished with the task or needs more input or action from the user in order to proceed.
If the assistant is finished with the task you MUST include <finish></finish> in your response.
IMPORTANT: Execute code using <execute_ipython>, <execute_bash>, or <execute_browse> whenever possible.
When handling files, try to use full paths and pwd to avoid errors.
"""
@@ -58,7 +60,9 @@ create_file('app.py')
USER:
OBSERVATION:
[File: /workspace/app.py (1 lines total)]
(this is the beginning of the file)
1|
(this is the end of the file)
[File app.py created.]
ASSISTANT:
@@ -74,11 +78,17 @@ def index():
if __name__ == '__main__':
app.run(port=5000)\"\"\"
edit_file('app.py', start=1, end=1, content=EDITED_CODE)
insert_content_at_line(
'app.py',
1,
EDITED_CODE,
)
</execute_ipython>
USER:
OBSERVATION:
(this is the beginning of the file)
1|from flask import Flask
2|app = Flask(__name__)
3|
@@ -89,7 +99,8 @@ OBSERVATION:
8|
9|if __name__ == '__main__':
10| app.run(port=5000)
[File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary.]
(this is the end of the file)
[File updated (edited at line 1). Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary.]
ASSISTANT:
I have created a Python file `app.py` that will display a list of numbers from 1 to 10 when you run it. Let me run the Python file for you:
@@ -199,6 +210,7 @@ open_file('app.py')
USER:
[File: /workspace/app.py (10 lines total)]
(this is the beginning of the file)
1|from flask import Flask
2|app = Flask(__name__)
3|
@@ -209,16 +221,22 @@ USER:
8|
9|if __name__ == '__main__':
10| app.run(port=5000)
(this is the end of the file)
ASSISTANT:
I should edit the file to display the numbers in a table format. I should include correct indentation. Let me update the file:
<execute_ipython>
edit_file('app.py', start=7, end=7, content=" return '<table>' + ''.join([f'<tr><td>{i}</td></tr>' for i in numbers]) + '</table>'")
edit_file_by_replace(
'app.py',
to_replace=" return str(numbers)",
new_content=" return '<table>' + ''.join([f'<tr><td>{i}</td></tr>' for i in numbers]) + '</table>'",
)
</execute_ipython>
USER:
Observation:
[File: /workspace/app.py (10 lines total after edit)]
(this is the beginning of the file)
1|from flask import Flask
2|app = Flask(__name__)
3|
@@ -229,7 +247,8 @@ Observation:
8|
9|if __name__ == '__main__':
10| app.run(port=5000)
[File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary.]
(this is the end of the file)
[File updated (edited at line 7). Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary.]
ASSISTANT:
Running the updated file:
@@ -243,6 +262,7 @@ Observation:
ASSISTANT:
The server is running on port 5000 with PID 126. You can access the list of numbers in a table format by visiting http://127.0.0.1:5000. Let me know if you have any further requests!
<finish></finish>
--- END OF EXAMPLE ---
"""
+2 -2
View File
@@ -2,6 +2,6 @@
This agent is an adaptation of the original [SWE Agent](https://swe-agent.com/) based on CodeAct using the `agentskills` library of OpenDevin.
Its intended use is **solving Github issues**.
Its intended use is **solving GitHub issues**.
It removes web-browsing and Github capability from the original CodeAct agent to avoid confusion to the agent.
It removes web-browsing and GitHub capability from the original CodeAct agent to avoid confusion to the agent.
+114
View File
@@ -0,0 +1,114 @@
import re
from opendevin.controller.action_parser import ActionParser
from opendevin.events.action import (
Action,
AgentFinishAction,
CmdRunAction,
IPythonRunCellAction,
MessageAction,
)
class CodeActSWEActionParserFinish(ActionParser):
"""
Parser action:
- AgentFinishAction() - end the interaction
"""
def __init__(
self,
):
self.finish_command = None
def check_condition(self, action_str: str) -> bool:
self.finish_command = re.search(r'<finish>.*</finish>', action_str, re.DOTALL)
return self.finish_command is not None
def parse(self, action_str: str) -> Action:
assert (
self.finish_command is not None
), 'self.finish_command should not be None when parse is called'
thought = action_str.replace(self.finish_command.group(0), '').strip()
return AgentFinishAction(thought=thought)
class CodeActSWEActionParserCmdRun(ActionParser):
"""
Parser action:
- CmdRunAction(command) - bash command to run
- AgentFinishAction() - end the interaction
"""
def __init__(
self,
):
self.bash_command = None
def check_condition(self, action_str: str) -> bool:
self.bash_command = re.search(
r'<execute_bash>(.*?)</execute_bash>', action_str, re.DOTALL
)
return self.bash_command is not None
def parse(self, action_str: str) -> Action:
assert (
self.bash_command is not None
), 'self.bash_command should not be None when parse is called'
thought = action_str.replace(self.bash_command.group(0), '').strip()
# a command was found
command_group = self.bash_command.group(1).strip()
if command_group.strip() == 'exit':
return AgentFinishAction()
return CmdRunAction(command=command_group, thought=thought)
class CodeActSWEActionParserIPythonRunCell(ActionParser):
"""
Parser action:
- IPythonRunCellAction(code) - IPython code to run
"""
def __init__(
self,
):
self.python_code = None
self.jupyter_kernel_init_code: str = 'from agentskills import *'
def check_condition(self, action_str: str) -> bool:
self.python_code = re.search(
r'<execute_ipython>(.*?)</execute_ipython>', action_str, re.DOTALL
)
return self.python_code is not None
def parse(self, action_str: str) -> Action:
assert (
self.python_code is not None
), 'self.python_code should not be None when parse is called'
code_group = self.python_code.group(1).strip()
thought = action_str.replace(self.python_code.group(0), '').strip()
return IPythonRunCellAction(
code=code_group,
thought=thought,
kernel_init_code=self.jupyter_kernel_init_code,
)
class CodeActSWEActionParserMessage(ActionParser):
"""
Parser action:
- MessageAction(content) - Message action to run (e.g. ask for clarification)
"""
def __init__(
self,
):
pass
def check_condition(self, action_str: str) -> bool:
# We assume the LLM is GOOD enough that when it returns pure natural language
# it wants to talk to the user
return True
def parse(self, action_str: str) -> Action:
return MessageAction(content=action_str, wait_for_response=True)
+54 -107
View File
@@ -1,26 +1,25 @@
import re
from agenthub.codeact_swe_agent.prompt import (
COMMAND_DOCS,
MINIMAL_SYSTEM_PREFIX,
SWE_EXAMPLE,
SYSTEM_PREFIX,
SYSTEM_SUFFIX,
)
from agenthub.codeact_swe_agent.response_parser import CodeActSWEResponseParser
from opendevin.controller.agent import Agent
from opendevin.controller.state.state import State
from opendevin.core.config import config
from opendevin.events.action import (
Action,
AgentFinishAction,
BrowseInteractiveAction,
CmdRunAction,
IPythonRunCellAction,
MessageAction,
)
from opendevin.events.observation import (
BrowserOutputObservation,
CmdOutputObservation,
IPythonRunCellObservation,
)
from opendevin.events.serialization.event import truncate_content
from opendevin.llm.llm import LLM
from opendevin.runtime.plugins import (
AgentSkillsRequirement,
@@ -30,21 +29,11 @@ from opendevin.runtime.plugins import (
from opendevin.runtime.tools import RuntimeTool
def parse_response(response) -> str:
action = response.choices[0].message.content
for lang in ['bash', 'ipython', 'browse']:
if f'<execute_{lang}>' in action and f'</execute_{lang}>' not in action:
action += f'</execute_{lang}>'
return action
def action_to_str(action: Action) -> str:
if isinstance(action, CmdRunAction):
return f'{action.thought}\n<execute_bash>\n{action.command}\n</execute_bash>'
elif isinstance(action, IPythonRunCellAction):
return f'{action.thought}\n<execute_ipython>\n{action.code}\n</execute_ipython>'
elif isinstance(action, BrowseInteractiveAction):
return f'{action.thought}\n<execute_browse>\n{action.browser_actions}\n</execute_browse>'
elif isinstance(action, MessageAction):
return action.content
return ''
@@ -52,8 +41,7 @@ def action_to_str(action: Action) -> str:
def get_action_message(action: Action) -> dict[str, str] | None:
if (
isinstance(action, BrowseInteractiveAction)
or isinstance(action, CmdRunAction)
isinstance(action, CmdRunAction)
or isinstance(action, IPythonRunCellAction)
or isinstance(action, MessageAction)
):
@@ -65,10 +53,13 @@ def get_action_message(action: Action) -> dict[str, str] | None:
def get_observation_message(obs) -> dict[str, str] | None:
max_message_chars = config.get_llm_config_from_agent(
'CodeActSWEAgent'
).max_message_chars
if isinstance(obs, CmdOutputObservation):
content = 'OBSERVATION:\n' + truncate_observation(obs.content)
content = 'OBSERVATION:\n' + truncate_content(obs.content, max_message_chars)
content += (
f'\n[Command {obs.command_id} finished with exit code {obs.exit_code}]]'
f'\n[Command {obs.command_id} finished with exit code {obs.exit_code}]'
)
return {'role': 'user', 'content': content}
elif isinstance(obs, IPythonRunCellObservation):
@@ -81,30 +72,13 @@ def get_observation_message(obs) -> dict[str, str] | None:
'![image](data:image/png;base64, ...) already displayed to user'
)
content = '\n'.join(splitted)
content = truncate_observation(content)
return {'role': 'user', 'content': content}
elif isinstance(obs, BrowserOutputObservation):
content = 'OBSERVATION:\n' + truncate_observation(obs.content)
content = truncate_content(content, max_message_chars)
return {'role': 'user', 'content': content}
return None
def truncate_observation(observation: str, max_chars: int = 10_000) -> str:
"""
Truncate the middle of the observation if it is too long.
"""
if len(observation) <= max_chars:
return observation
half = max_chars // 2
return (
observation[:half]
+ '\n[... Observation truncated due to length ...]\n'
+ observation[-half:]
)
def get_system_message() -> str:
return f'{MINIMAL_SYSTEM_PREFIX}\n\n{COMMAND_DOCS}\n\n{SYSTEM_SUFFIX}'
return f'{SYSTEM_PREFIX}\n\n{COMMAND_DOCS}\n\n{SYSTEM_SUFFIX}'
def get_in_context_example() -> str:
@@ -112,7 +86,7 @@ def get_in_context_example() -> str:
class CodeActSWEAgent(Agent):
VERSION = '1.5'
VERSION = '1.6'
"""
This agent is an adaptation of the original [SWE Agent](https://swe-agent.com/) based on CodeAct 1.5 using the `agentskills` library of OpenDevin.
@@ -123,17 +97,18 @@ class CodeActSWEAgent(Agent):
sandbox_plugins: list[PluginRequirement] = [
# NOTE: AgentSkillsRequirement need to go before JupyterRequirement, since
# AgentSkillsRequirement provides a lot of Python functions
# and it need to be initialized before Jupyter for Jupyter to use those functions.
# AgentSkillsRequirement provides a lot of Python functions,
# and it needs to be initialized before Jupyter for Jupyter to use those functions.
AgentSkillsRequirement(),
JupyterRequirement(),
]
runtime_tools: list[RuntimeTool] = [RuntimeTool.BROWSER]
jupyter_kernel_init_code: str = 'from agentskills import *'
runtime_tools: list[RuntimeTool] = []
system_message: str = get_system_message()
in_context_example: str = f"Here is an example of how you can interact with the environment for task solving:\n{get_in_context_example()}\n\nNOW, LET'S START!"
response_parser = CodeActSWEResponseParser()
def __init__(
self,
llm: LLM,
@@ -164,85 +139,57 @@ class CodeActSWEAgent(Agent):
Returns:
- CmdRunAction(command) - bash command to run
- IPythonRunCellAction(code) - IPython code to run
- BrowseInteractiveAction(browsergym_command) - BrowserGym commands to run
- MessageAction(content) - Message action to run (e.g. ask for clarification)
- AgentFinishAction() - end the interaction
"""
messages: list[dict[str, str]] = [
{'role': 'system', 'content': self.system_message},
{'role': 'user', 'content': self.in_context_example},
]
for prev_action, obs in state.history:
action_message = get_action_message(prev_action)
if action_message:
messages.append(action_message)
# if we're done, go back
latest_user_message = state.history.get_last_user_message()
if latest_user_message and latest_user_message.strip() == '/exit':
return AgentFinishAction()
obs_message = get_observation_message(obs)
if obs_message:
messages.append(obs_message)
# prepare what we want to send to the LLM
messages: list[dict[str, str]] = self._get_messages(state)
latest_user_message = [m for m in messages if m['role'] == 'user'][-1]
if latest_user_message:
if latest_user_message['content'].strip() == '/exit':
return AgentFinishAction()
latest_user_message['content'] += (
f'\n\nENVIRONMENT REMINDER: You have {state.max_iterations - state.iteration} turns left to complete the task.'
)
response = self.llm.do_completion(
response = self.llm.completion(
messages=messages,
stop=[
'</execute_ipython>',
'</execute_bash>',
'</execute_browse>',
],
temperature=0.0,
)
action_str: str = parse_response(response)
state.num_of_chars += sum(
len(message['content']) for message in messages
) + len(action_str)
return self.response_parser.parse(response)
if finish_command := re.search(r'<finish>.*</finish>', action_str, re.DOTALL):
thought = action_str.replace(finish_command.group(0), '').strip()
return AgentFinishAction(thought=thought)
if bash_command := re.search(
r'<execute_bash>(.*?)</execute_bash>', action_str, re.DOTALL
):
# remove the command from the action string to get thought
thought = action_str.replace(bash_command.group(0), '').strip()
# a command was found
command_group = bash_command.group(1).strip()
def _get_messages(self, state: State) -> list[dict[str, str]]:
messages = [
{'role': 'system', 'content': self.system_message},
{'role': 'user', 'content': self.in_context_example},
]
if command_group.strip() == 'exit':
return AgentFinishAction()
return CmdRunAction(command=command_group, thought=thought)
elif python_code := re.search(
r'<execute_ipython>(.*?)</execute_ipython>', action_str, re.DOTALL
):
# a code block was found
code_group = python_code.group(1).strip()
thought = action_str.replace(python_code.group(0), '').strip()
return IPythonRunCellAction(
code=code_group,
thought=thought,
kernel_init_code=self.jupyter_kernel_init_code,
for event in state.history.get_events():
# create a regular message from an event
message = (
get_action_message(event)
if isinstance(event, Action)
else get_observation_message(event)
)
elif browse_command := re.search(
r'<execute_browse>(.*)</execute_browse>', action_str, re.DOTALL
):
# BrowserGym actions was found
browse_actions = browse_command.group(1).strip()
thought = action_str.replace(browse_command.group(0), '').strip()
return BrowseInteractiveAction(
browser_actions=browse_actions, thought=thought
)
else:
# We assume the LLM is GOOD enough that when it returns pure natural language
# it want to talk to the user
return MessageAction(content=action_str, wait_for_response=True)
def search_memory(self, query: str) -> list[str]:
raise NotImplementedError('Implement this abstract method')
# add regular message
if message:
messages.append(message)
# the latest user message is important:
# we want to remind the agent of the environment constraints
latest_user_message = next(
(m for m in reversed(messages) if m['role'] == 'user'), None
)
# add a reminder to the prompt
if latest_user_message:
latest_user_message['content'] += (
f'\n\nENVIRONMENT REMINDER: You have {state.max_iterations - state.iteration} turns left to complete the task.'
)
return messages
+6 -2
View File
@@ -18,9 +18,13 @@ The assistant can execute bash commands on behalf of the user by wrapping them w
For example, you can list the files in the current directory by <execute_bash> ls </execute_bash>.
"""
PIP_INSTALL_PREFIX = """The assistant can install Python packages using the %pip magic command in an IPython environment by using the following syntax: <execute_ipython> %pip install [package needed] </execute_ipython> and should always import packages and define variables before starting to use them."""
SYSTEM_PREFIX = MINIMAL_SYSTEM_PREFIX + PIP_INSTALL_PREFIX
SYSTEM_SUFFIX = """The assistant's response should be concise.
The assistant should include ONLY ONE <execute_ipython> or <execute_bash> or <execute_browse> in every one of the responses, unless the assistant is finished with the task or need more input or action from the user in order to proceed.
IMPORTANT: Whenever possible, execute the code for the user using <execute_ipython> or <execute_bash> or <execute_browse> instead of providing it.
The assistant should include ONLY ONE <execute_ipython> or <execute_bash> in every one of the responses, unless the assistant is finished with the task or need more input or action from the user in order to proceed.
IMPORTANT: Whenever possible, execute the code for the user using <execute_ipython> or <execute_bash> instead of providing it.
"""
SWE_EXAMPLE = """
@@ -0,0 +1,47 @@
from agenthub.codeact_swe_agent.action_parser import (
CodeActSWEActionParserCmdRun,
CodeActSWEActionParserFinish,
CodeActSWEActionParserIPythonRunCell,
CodeActSWEActionParserMessage,
)
from opendevin.controller.action_parser import ResponseParser
from opendevin.events.action import Action
class CodeActSWEResponseParser(ResponseParser):
"""
Parser action:
- CmdRunAction(command) - bash command to run
- IPythonRunCellAction(code) - IPython code to run
- MessageAction(content) - Message action to run (e.g. ask for clarification)
- AgentFinishAction() - end the interaction
"""
def __init__(self):
# Need pay attention to the item order in self.action_parsers
super().__init__()
self.action_parsers = [
CodeActSWEActionParserFinish(),
CodeActSWEActionParserCmdRun(),
CodeActSWEActionParserIPythonRunCell(),
]
self.default_parser = CodeActSWEActionParserMessage()
def parse(self, response: str) -> Action:
action_str = self.parse_response(response)
return self.parse_action(action_str)
def parse_response(self, response) -> str:
action = response.choices[0].message.content
if action is None:
return ''
for lang in ['bash', 'ipython']:
if f'<execute_{lang}>' in action and f'</execute_{lang}>' not in action:
action += f'</execute_{lang}>'
return action
def parse_action(self, action_str: str) -> Action:
for action_parser in self.action_parsers:
if action_parser.check_condition(action_str):
return action_parser.parse(action_str)
return self.default_parser.parse(action_str)
+6 -8
View File
@@ -8,8 +8,7 @@ from opendevin.llm.llm import LLM
class DelegatorAgent(Agent):
VERSION = '1.0'
"""
The planner agent utilizes a special prompting strategy to create long term plans for solving problems.
The agent is given its previous action-observation pairs, current task, and hint based on last action taken at every step.
The Delegator Agent is responsible for delegating tasks to other agents based on the current task.
"""
current_delegate: str = ''
@@ -26,14 +25,14 @@ class DelegatorAgent(Agent):
def step(self, state: State) -> Action:
"""
Checks to see if current step is completed, returns AgentFinishAction if True.
Otherwise, creates a plan prompt and sends to model for inference, returning the result as the next action.
Otherwise, delegates the task to the next agent in the pipeline.
Parameters:
- state (State): The current state given the previous actions and observations
Returns:
- AgentFinishAction: If the last state was 'completed', 'verified', or 'abandoned'
- Action: The next action to take based on llm response
- AgentDelegateAction: The next agent to delegate the task to
"""
if self.current_delegate == '':
self.current_delegate = 'study'
@@ -42,7 +41,9 @@ class DelegatorAgent(Agent):
agent='StudyRepoForTaskAgent', inputs={'task': task}
)
last_observation = state.history[-1][1]
# last observation in history should be from the delegate
last_observation = state.history.get_last_observation()
if not isinstance(last_observation, AgentDelegateObservation):
raise Exception('Last observation is not an AgentDelegateObservation')
@@ -81,6 +82,3 @@ class DelegatorAgent(Agent):
)
else:
raise Exception('Invalid delegate state')
def search_memory(self, query: str) -> list[str]:
return []
+7 -36
View File
@@ -7,7 +7,6 @@ from opendevin.events.action import (
Action,
AddTaskAction,
AgentFinishAction,
AgentRecallAction,
AgentRejectAction,
BrowseInteractiveAction,
BrowseURLAction,
@@ -18,7 +17,6 @@ from opendevin.events.action import (
ModifyTaskAction,
)
from opendevin.events.observation import (
AgentRecallObservation,
CmdOutputObservation,
FileReadObservation,
FileWriteObservation,
@@ -31,8 +29,6 @@ from opendevin.llm.llm import LLM
"""
FIXME: There are a few problems this surfaced
* FileWrites seem to add an unintended newline at the end of the file
* command_id is sometimes a number, sometimes a string
* Why isn't the output of the background command split between two steps?
* Browser not working
"""
@@ -40,8 +36,6 @@ ActionObs = TypedDict(
'ActionObs', {'action': Action, 'observations': list[Observation]}
)
BACKGROUND_CMD = 'echo "This is in the background" && sleep .1 && echo "This too"'
class DummyAgent(Agent):
VERSION = '1.0'
@@ -95,28 +89,6 @@ class DummyAgent(Agent):
)
],
},
{
'action': CmdRunAction(command=BACKGROUND_CMD, background=True),
'observations': [
CmdOutputObservation(
'Background command started. To stop it, send a `kill` action with command_id 42',
command_id='42', # type: ignore[arg-type]
command=BACKGROUND_CMD,
),
CmdOutputObservation(
'This is in the background\nThis too\n',
command_id='42', # type: ignore[arg-type]
command=BACKGROUND_CMD,
),
],
},
{
'action': AgentRecallAction(query='who am I?'),
'observations': [
AgentRecallObservation('', memories=['I am a computer.']),
# CmdOutputObservation('This too\n', command_id='42', command=BACKGROUND_CMD),
],
},
{
'action': BrowseURLAction(url='https://google.com'),
'observations': [
@@ -145,11 +117,16 @@ class DummyAgent(Agent):
time.sleep(0.1)
if state.iteration > 0:
prev_step = self.steps[state.iteration - 1]
# a step is (action, observations list)
if 'observations' in prev_step:
# one obs, at most
expected_observations = prev_step['observations']
hist_start = len(state.history) - len(expected_observations)
# check if the history matches the expected observations
hist_events = state.history.get_last_events(len(expected_observations))
for i in range(len(expected_observations)):
hist_obs = event_to_dict(state.history[hist_start + i][1])
hist_obs = event_to_dict(hist_events[i])
expected_obs = event_to_dict(expected_observations[i])
if (
'command_id' in hist_obs['extras']
@@ -163,13 +140,7 @@ class DummyAgent(Agent):
):
del expected_obs['extras']['command_id']
expected_obs['content'] = ''
if hist_obs != expected_obs:
print('\nactual', hist_obs)
print('\nexpect', expected_obs)
assert (
hist_obs == expected_obs
), f'Expected observation {expected_obs}, got {hist_obs}'
return self.steps[state.iteration]['action']
def search_memory(self, query: str) -> list[str]:
return ['I am a computer.']
@@ -1,2 +0,0 @@
* `kill` - kills a background command
* `command_id` - the ID of the background command to kill
@@ -1,3 +1,2 @@
* `run` - runs a command on the command line in a Linux shell. Arguments:
* `command` - the command to run
* `background` - if true, run the command in the background, so that other commands can be run concurrently. Useful for e.g. starting a server. You won't be able to see the logs. You don't need to end the command with `&`, just set this to true.
+21 -14
View File
@@ -2,11 +2,13 @@ from jinja2 import BaseLoader, Environment
from opendevin.controller.agent import Agent
from opendevin.controller.state.state import State
from opendevin.core.config import config
from opendevin.core.utils import json
from opendevin.events.action import Action
from opendevin.events.serialization.action import action_from_dict
from opendevin.events.serialization.event import event_to_memory
from opendevin.llm.llm import LLM
from opendevin.memory.history import ShortTermHistory
from .instructions import instructions
from .registry import all_microagents
@@ -27,18 +29,27 @@ def to_json(obj, **kwargs):
return json.dumps(obj, **kwargs)
def history_to_json(obj, **kwargs):
def history_to_json(history: ShortTermHistory, max_events=20, **kwargs):
"""
Serialize and simplify history to str format
"""
if isinstance(obj, list):
# process history, make it simpler.
processed_history = []
for action, observation in obj:
processed_history.append(
(event_to_memory(action), event_to_memory(observation))
)
return json.dumps(processed_history, **kwargs)
# TODO: get agent specific llm config
llm_config = config.get_llm_config()
max_message_chars = llm_config.max_message_chars
processed_history = []
event_count = 0
for event in history.get_events(reverse=True):
if event_count >= max_events:
break
processed_history.append(event_to_memory(event, max_message_chars))
event_count += 1
# history is in reverse order, let's fix it
processed_history.reverse()
return json.dumps(processed_history, **kwargs)
class MicroAgent(Agent):
@@ -64,11 +75,7 @@ class MicroAgent(Agent):
latest_user_message=state.get_current_user_intent(),
)
messages = [{'content': prompt, 'role': 'user'}]
resp = self.llm.do_completion(messages=messages)
resp = self.llm.completion(messages=messages)
action_resp = resp['choices'][0]['message']['content']
state.num_of_chars += len(prompt) + len(action_resp)
action = parse_response(action_resp)
return action
def search_memory(self, query: str) -> list[str]:
return []
+1 -1
View File
@@ -21,7 +21,7 @@ Do NOT finish until you have completed the tasks.
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
## Format
{{ instructions.format.action }}
+1 -1
View File
@@ -3,7 +3,7 @@
CommitWriterAgent can help write git commit message. Example:
```bash
WORKSPACE_MOUNT_PATH="`PWD`" SANDBOX_TYPE="exec" \
WORKSPACE_MOUNT_PATH="`PWD`" SANDBOX_BOX_TYPE="ssh" \
poetry run python opendevin/core/main.py -t "dummy task" -c CommitWriterAgent -d ./
```
+7 -5
View File
@@ -11,14 +11,16 @@ changes. The commit message should include:
- A summary line that clearly states the purpose of the changes.
- Optionally, a detailed description if the changes are complex or need further explanation.
You should find the diff using `git diff --cached`, compile a commit message,
and call the `finish` action with `outputs.answer` set to the answer. If current
repo is not a valid git repo, or there is no diff in the staging area, please call
the `reject` action.
You should first use `git status` to check whether it's a valid git repo and there
is diff in the staging area. If not, please call the `reject` action.
If it is a valid git repo and there is diff in the staging area, you should find
the diff using `git diff --cached`, compile a commit message, and call the `finish`
action with `outputs.answer` set to the answer.
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
If the last item in the history is an error, you should try to fix it.
+3 -2
View File
@@ -14,7 +14,8 @@ continue to delegate to one of the agents below to until the full task is finish
task you send to. You should revisit the input you send to the delegate, and consider
whether any other delegate would be able to solve the task. If you cannot find
a proper delegate agent, or the delegate attempts keep failing, call the `reject`
action.
action. In `reason` attribute, make sure you include your attempts (e.g. what agent
you have delegated to, and why they failed).
## Agents
{% for name, details in delegates.items() %}
@@ -26,7 +27,7 @@ action.
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
If the last item in the history is an error, you should try to fix it. If you
cannot fix it, call the `reject` action.
+1 -1
View File
@@ -10,7 +10,7 @@ and call the `finish` action with `outputs.answer` set to the answer.
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
If the last item in the history is an error, you should try to fix it.
+1 -1
View File
@@ -18,7 +18,7 @@ You may take any of the following actions:
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
## Format
{{ instructions.format.action }}
+1 -1
View File
@@ -20,7 +20,7 @@ When you're done, put your summary into the output of the `finish` action.
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
## Format
{{ instructions.format.action }}
+2 -3
View File
@@ -24,7 +24,7 @@ implement the solution. If the codebase is empty, you should call the `finish` a
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
## Format
{{ instructions.format.action }}
@@ -41,8 +41,7 @@ ASSISTANT:
{
"action": "run",
"args": {
"command": "ls",
"background": false
"command": "ls"
}
}
+1 -1
View File
@@ -31,7 +31,7 @@ Do NOT finish until you have fixed all the typos and generated a summary.
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-5:]) }}
{{ history_to_json(state.history, max_events=10) }}
## Format
{{ instructions.format.action }}
+1 -1
View File
@@ -22,7 +22,7 @@ explaining what the problem is.
## History
{{ instructions.history_truncated }}
{{ history_to_json(state.history[-10:]) }}
{{ history_to_json(state.history, max_events=20) }}
## Format
{{ instructions.format.action }}
+32 -81
View File
@@ -1,4 +1,5 @@
import agenthub.monologue_agent.utils.prompts as prompts
from agenthub.monologue_agent.response_parser import MonologueResponseParser
from agenthub.monologue_agent.utils.prompts import INITIAL_THOUGHTS
from opendevin.controller.agent import Agent
from opendevin.controller.state.state import State
@@ -7,7 +8,6 @@ from opendevin.core.exceptions import AgentNoInstructionError
from opendevin.core.schema import ActionType
from opendevin.events.action import (
Action,
AgentRecallAction,
BrowseURLAction,
CmdRunAction,
FileReadAction,
@@ -16,7 +16,6 @@ from opendevin.events.action import (
NullAction,
)
from opendevin.events.observation import (
AgentRecallObservation,
BrowserOutputObservation,
CmdOutputObservation,
FileReadObservation,
@@ -28,12 +27,9 @@ from opendevin.llm.llm import LLM
from opendevin.memory.condenser import MemoryCondenser
from opendevin.runtime.tools import RuntimeTool
if config.agent.memory_enabled:
if config.get_agent_config('MonologueAgent').memory_enabled:
from opendevin.memory.memory import LongTermMemory
MAX_TOKEN_COUNT_PADDING = 512
MAX_OUTPUT_LENGTH = 5000
class MonologueAgent(Agent):
VERSION = '1.0'
@@ -48,6 +44,7 @@ class MonologueAgent(Agent):
memory: 'LongTermMemory | None'
memory_condenser: MemoryCondenser
runtime_tools: list[RuntimeTool] = [RuntimeTool.BROWSER]
response_parser = MonologueResponseParser()
def __init__(self, llm: LLM):
"""
@@ -66,7 +63,7 @@ class MonologueAgent(Agent):
Will execute again when called after reset.
Parameters:
- task (str): The initial goal statement provided by the user
- task: The initial goal statement provided by the user
Raises:
- AgentNoInstructionError: If task is not provided
@@ -79,7 +76,7 @@ class MonologueAgent(Agent):
raise AgentNoInstructionError()
self.initial_thoughts = []
if config.agent.memory_enabled:
if config.get_agent_config('MonologueAgent').memory_enabled:
self.memory = LongTermMemory()
else:
self.memory = None
@@ -90,6 +87,9 @@ class MonologueAgent(Agent):
self._initialized = True
def _add_initial_thoughts(self, task):
max_message_chars = config.get_llm_config_from_agent(
'MonologueAgent'
).max_message_chars
previous_action = ''
for thought in INITIAL_THOUGHTS:
thought = thought.replace('$TASK', task)
@@ -101,13 +101,13 @@ class MonologueAgent(Agent):
)
elif previous_action == ActionType.READ:
observation = FileReadObservation(content=thought, path='')
elif previous_action == ActionType.RECALL:
observation = AgentRecallObservation(content=thought, memories=[])
elif previous_action == ActionType.BROWSE:
observation = BrowserOutputObservation(
content=thought, url='', screenshot=''
)
self.initial_thoughts.append(event_to_memory(observation))
self.initial_thoughts.append(
event_to_memory(observation, max_message_chars)
)
previous_action = ''
else:
action: Action = NullAction()
@@ -124,17 +124,13 @@ class MonologueAgent(Agent):
path = thought.split('READ ')[1]
action = FileReadAction(path=path)
previous_action = ActionType.READ
elif thought.startswith('RECALL'):
query = thought.split('RECALL ')[1]
action = AgentRecallAction(query=query)
previous_action = ActionType.RECALL
elif thought.startswith('BROWSE'):
url = thought.split('BROWSE ')[1]
action = BrowseURLAction(url=url)
previous_action = ActionType.BROWSE
else:
action = MessageAction(thought)
self.initial_thoughts.append(event_to_memory(action))
self.initial_thoughts.append(event_to_memory(action, max_message_chars))
def step(self, state: State) -> Action:
"""
@@ -146,32 +142,35 @@ class MonologueAgent(Agent):
Returns:
- Action: The next action to take based on LLM response
"""
max_message_chars = config.get_llm_config_from_agent(
'MonologueAgent'
).max_message_chars
goal = state.get_current_user_intent()
self._initialize(goal)
recent_events: list[dict[str, str]] = []
# add the events from state.history
for prev_action, obs in state.history:
if not isinstance(prev_action, NullAction):
recent_events.append(event_to_memory(prev_action))
if not isinstance(obs, NullObservation):
recent_events.append(self._truncate_output(event_to_memory(obs)))
for event in state.history.get_events():
recent_events.append(event_to_memory(event, max_message_chars))
# add the last messages to long term memory
if self.memory is not None and state.history and len(state.history) > 0:
self.memory.add_event(event_to_memory(state.history[-1][0]))
self.memory.add_event(
self._truncate_output(event_to_memory(state.history[-1][1]))
)
if self.memory is not None:
last_action = state.history.get_last_action()
last_observation = state.history.get_last_observation()
# this should still work
# we will need to do this differently: find out if there really is an action or an observation in this step
if last_action:
self.memory.add_event(event_to_memory(last_action, max_message_chars))
if last_observation:
self.memory.add_event(
event_to_memory(last_observation, max_message_chars)
)
# the action prompt with initial thoughts and recent events
prompt = prompts.get_request_action_prompt(
goal,
self.initial_thoughts,
recent_events,
state.background_commands_obs,
goal, self.initial_thoughts, recent_events
)
messages: list[dict[str, str]] = [
@@ -179,60 +178,12 @@ class MonologueAgent(Agent):
]
# format all as a single message, a monologue
resp = self.llm.do_completion(messages=messages)
resp = self.llm.completion(messages=messages)
# get the next action from the response
action_resp = resp['choices'][0]['message']['content']
# keep track of max_chars fallback option
state.num_of_chars += len(prompt) + len(action_resp)
action = prompts.parse_action_response(action_resp)
action = self.response_parser.parse(resp)
self.latest_action = action
return action
def _truncate_output(
self, observation: dict, max_chars: int = MAX_OUTPUT_LENGTH
) -> dict[str, str]:
"""
Truncates the output of an observation to a maximum number of characters.
Parameters:
- output (str): The observation whose output to truncate
- max_chars (int): The maximum number of characters to allow
Returns:
- str: The truncated output
"""
if (
'args' in observation
and 'output' in observation['args']
and len(observation['args']['output']) > max_chars
):
output = observation['args']['output']
half = max_chars // 2
observation['args']['output'] = (
output[:half]
+ '\n[... Output truncated due to length...]\n'
+ output[-half:]
)
return observation
def search_memory(self, query: str) -> list[str]:
"""
Uses VectorIndexRetriever to find related memories within the long term memory.
Uses search to produce top 10 results.
Parameters:
- query (str): The query that we want to find related memories for
Returns:
- list[str]: A list of top 10 text results that matched the query
"""
if self.memory is None:
return []
return self.memory.search(query)
def reset(self) -> None:
super().reset()
@@ -0,0 +1,38 @@
from opendevin.controller.action_parser import ResponseParser
from opendevin.core.utils import json
from opendevin.events.action import (
Action,
)
from opendevin.events.serialization.action import action_from_dict
class MonologueResponseParser(ResponseParser):
def __init__(self):
super().__init__()
def parse(self, response: str) -> Action:
action_str = self.parse_response(response)
return self.parse_action(action_str)
def parse_response(self, response) -> str:
# get the next action from the response
return response['choices'][0]['message']['content']
def parse_action(self, action_str: str) -> Action:
"""
Parses a string to find an action within it
Parameters:
- response (str): The string to be parsed
Returns:
- Action: The action that was found in the response string
"""
# attempt to load the JSON dict from the response
action_dict = json.loads(action_str)
if 'content' in action_dict:
# The LLM gets confused here. Might as well be robust
action_dict['contents'] = action_dict.pop('content')
return action_from_dict(action_dict)
+3 -36
View File
@@ -3,9 +3,6 @@ from opendevin.core.utils import json
from opendevin.events.action import (
Action,
)
from opendevin.events.observation import (
CmdOutputObservation,
)
from opendevin.events.serialization.action import action_from_dict
ACTION_PROMPT = """
@@ -32,25 +29,18 @@ Here are the possible actions:
* `content` - the content to write to the file
* `run` - runs a command. Arguments:
* `command` - the command to run
* `background` - if true, run the command in the background, so that other commands can be run concurrently. Useful for e.g. starting a server. You won't be able to see the logs. You don't need to end the command with `&`, just set this to true.
* `kill` - kills a background command
* `command_id` - the ID of the background command to kill
* `browse` - opens a web page. Arguments:
* `url` - the URL to open
* `push` - Push a branch from the current repo to github:
* `owner` - the owner of the repo to push to
* `repo` - the name of the repo to push to
* `branch` - the name of the branch to push
* `recall` - recalls a past memory. Arguments:
* `query` - the query to search for
* `message` - make a plan, set a goal, record your thoughts, or ask for more input from the user. Arguments:
* `content` - the message to record
* `wait_for_response` - set to `true` to wait for the user to respond before proceeding
* `finish` - if you're absolutely certain that you've completed your task and have tested your work, use the finish action to stop working.
%(background_commands)s
You MUST take time to think in between read, write, run, kill, browse, push, and recall actions--do this with the `message` action.
You MUST take time to think in between read, write, run, browse, and push actions--do this with the `message` action.
You should never act twice in a row without thinking. But if your last several
actions are all `message` actions, you should consider taking a different action.
@@ -100,15 +90,7 @@ INITIAL_THOUGHTS = [
'It seems like I have some kind of short term memory.',
'Each of my thoughts seems to be stored in a JSON array.',
'It seems whatever I say next will be added as an object to the list.',
'But no one has perfect short-term memory. My list of thoughts will be summarized and condensed over time, losing information in the process.',
'Fortunately I have long term memory!',
'I can just perform a recall action, followed by the thing I want to remember. And then related thoughts just spill out!',
"Sometimes they're random thoughts that don't really have to do with what I wanted to remember. But usually they're exactly what I need!",
"Let's try it out!",
'RECALL what it is I want to do',
"Here's what I want to do: $TASK",
'How am I going to get there though?',
"Neat! And it looks like it's easy for me to use the command line too! I just have to perform a run action and include the command I want to run in the command argument. The command output just jumps into my head!",
"It looks like it's easy for me to use the command line too! I just have to perform a run action and include the command I want to run in the command argument. The command output just jumps into my head!",
'RUN echo "hello world"',
'hello world',
'Cool! I bet I can write files too using the write action.',
@@ -153,7 +135,6 @@ def get_request_action_prompt(
task: str,
thoughts: list[dict],
recent_events: list[dict],
background_commands_obs: list[CmdOutputObservation] | None = None,
):
"""
Gets the action prompt formatted with appropriate values.
@@ -161,15 +142,11 @@ def get_request_action_prompt(
Parameters:
- task (str): The current task the agent is trying to accomplish
- thoughts (list[dict]): The agent's current thoughts
- background_commands_obs (list[CmdOutputObservation]): list of all observed background commands running
Returns:
- str: Formatted prompt string with hint, task, monologue, and background commands included
"""
if background_commands_obs is None:
background_commands_obs = []
hint = ''
if len(recent_events) > 0:
latest_event = recent_events[-1]
@@ -187,15 +164,6 @@ def get_request_action_prompt(
else:
hint = "You're just getting started! What should you do first?"
bg_commands_message = ''
if len(background_commands_obs) > 0:
bg_commands_message = 'The following commands are running in the background:'
for command_obs in background_commands_obs:
bg_commands_message += (
f'\n`{command_obs.command_id}`: {command_obs.command}'
)
bg_commands_message += '\nYou can end any process by sending a `kill` action with the numerical `command_id` above.'
user = 'opendevin' if config.run_as_devin else 'root'
monologue = thoughts + recent_events
@@ -203,10 +171,9 @@ def get_request_action_prompt(
return ACTION_PROMPT % {
'task': task,
'monologue': json.dumps(monologue, indent=2),
'background_commands': bg_commands_message,
'hint': hint,
'user': user,
'timeout': config.sandbox_timeout,
'timeout': config.sandbox.timeout,
'WORKSPACE_MOUNT_PATH_IN_SANDBOX': config.workspace_mount_path_in_sandbox,
}
+5 -9
View File
@@ -1,10 +1,11 @@
from agenthub.monologue_agent.response_parser import MonologueResponseParser
from opendevin.controller.agent import Agent
from opendevin.controller.state.state import State
from opendevin.events.action import Action, AgentFinishAction
from opendevin.llm.llm import LLM
from opendevin.runtime.tools import RuntimeTool
from .prompt import get_prompt, parse_response
from .prompt import get_prompt
class PlannerAgent(Agent):
@@ -14,6 +15,7 @@ class PlannerAgent(Agent):
The agent is given its previous action-observation pairs, current task, and hint based on last action taken at every step.
"""
runtime_tools: list[RuntimeTool] = [RuntimeTool.BROWSER]
response_parser = MonologueResponseParser()
def __init__(self, llm: LLM):
"""
@@ -45,11 +47,5 @@ class PlannerAgent(Agent):
return AgentFinishAction()
prompt = get_prompt(state)
messages = [{'content': prompt, 'role': 'user'}]
resp = self.llm.do_completion(messages=messages)
action_resp = resp['choices'][0]['message']['content']
state.num_of_chars += len(prompt) + len(action_resp)
action = parse_response(action_resp)
return action
def search_memory(self, query: str) -> list[str]:
return []
resp = self.llm.completion(messages=messages)
return self.response_parser.parse(resp)
+31 -20
View File
@@ -1,4 +1,5 @@
from opendevin.controller.state.state import State
from opendevin.core.config import config
from opendevin.core.logger import opendevin_logger as logger
from opendevin.core.schema import ActionType
from opendevin.core.utils import json
@@ -6,13 +7,10 @@ from opendevin.events.action import (
Action,
NullAction,
)
from opendevin.events.observation import (
NullObservation,
)
from opendevin.events.serialization.action import action_from_dict
from opendevin.events.serialization.event import event_to_memory
HISTORY_SIZE = 10
HISTORY_SIZE = 20
prompt = """
# Task
@@ -77,9 +75,6 @@ It must be an object, and it must contain two fields:
* `content` - the content to write to the file
* `run` - runs a command on the command line in a Linux shell. Arguments:
* `command` - the command to run
* `background` - if true, run the command in the background, so that other commands can be run concurrently. Useful for e.g. starting a server. You won't be able to see the logs. You don't need to end the command with `&`, just set this to true.
* `kill` - kills a background command
* `command_id` - the ID of the background command to kill
* `browse` - opens a web page. Arguments:
* `url` - the URL to open
* `message` - make a plan, set a goal, record your thoughts, or ask for more input from the user. Arguments:
@@ -94,7 +89,7 @@ It must be an object, and it must contain two fields:
* `state` - set to 'in_progress' to start the task, 'completed' to finish it, 'verified' to assert that it was successful, 'abandoned' to give up on it permanently, or `open` to stop working on it for now.
* `finish` - if ALL of your tasks and subtasks have been verified or abandoned, and you're absolutely certain that you've completed your task and have tested your work, use the finish action to stop working.
You MUST take time to think in between read, write, run, kill, browse, and recall actions--do this with the `message` action.
You MUST take time to think in between read, write, run, and browse actions--do this with the `message` action.
You should never act twice in a row without thinking. But if your last several
actions are all `message` actions, you should consider taking a different action.
@@ -114,7 +109,6 @@ def get_hint(latest_action_id: str) -> str:
ActionType.WRITE: 'You just changed a file. You should think about how it affects your plan.',
ActionType.BROWSE: 'You should think about the page you just visited, and what you learned from it.',
ActionType.MESSAGE: "Look at your last thought in the history above. What does it suggest? Don't think anymore--take action.",
ActionType.RECALL: 'You should think about the information you just recalled, and how it should affect your plan.',
ActionType.ADD_TASK: 'You should think about the next action to take.',
ActionType.MODIFY_TASK: 'You should think about the next action to take.',
ActionType.SUMMARIZE: '',
@@ -134,19 +128,32 @@ def get_prompt(state: State) -> str:
Returns:
- str: The formatted string prompt with historical values
"""
max_message_chars = config.get_llm_config_from_agent(
'PlannerAgent'
).max_message_chars
# the plan
plan_str = json.dumps(state.root_task.to_dict(), indent=2)
sub_history = state.history[-HISTORY_SIZE:]
# the history
history_dicts = []
latest_action: Action = NullAction()
for action, observation in sub_history:
if not isinstance(action, NullAction):
history_dicts.append(event_to_memory(action))
latest_action = action
if not isinstance(observation, NullObservation):
observation_dict = event_to_memory(observation)
history_dicts.append(observation_dict)
# retrieve the latest HISTORY_SIZE events
for event_count, event in enumerate(state.history.get_events(reverse=True)):
if event_count >= HISTORY_SIZE:
break
if latest_action == NullAction() and isinstance(event, Action):
latest_action = event
history_dicts.append(event_to_memory(event, max_message_chars))
# history_dicts is in reverse order, lets fix it
history_dicts.reverse()
# and get it as a JSON string
history_str = json.dumps(history_dicts, indent=2)
# the plan status
current_task = state.root_task.get_current_task()
if current_task is not None:
plan_status = f"You're currently working on this task:\n{current_task.goal}."
@@ -154,9 +161,15 @@ def get_prompt(state: State) -> str:
plan_status += "\nIf it's not achievable AND verifiable with a SINGLE action, you MUST break it down into subtasks NOW."
else:
plan_status = "You're not currently working on any tasks. Your next action MUST be to mark a task as in_progress."
hint = get_hint(event_to_memory(latest_action).get('action', ''))
# the hint, based on the last action
hint = get_hint(event_to_memory(latest_action, max_message_chars).get('action', ''))
logger.info('HINT:\n' + hint, extra={'msg_type': 'DETAIL'})
# the last relevant user message (the task)
task = state.get_current_user_intent()
# finally, fill in the prompt
return prompt % {
'task': task,
'plan': plan_str,
@@ -169,10 +182,8 @@ def get_prompt(state: State) -> str:
def parse_response(response: str) -> Action:
"""
Parses the model output to find a valid action to take
Parameters:
- response (str): A response from the model that potentially contains an Action.
Returns:
- Action: A valid next action to perform from model output
"""
+203
View File
@@ -0,0 +1,203 @@
###################### OpenDevin Configuration Example ######################
#
# All settings have default values, so you only need to uncomment and
# modify what you want to change
# The fields within each section are sorted in alphabetical order.
#
##############################################################################
#################################### Core ####################################
# General core configurations
##############################################################################
[core]
# API key for E2B
#e2b_api_key = ""
# Base path for the workspace
workspace_base = "./workspace"
# Cache directory path
#cache_dir = "/tmp/cache"
# Debugging enabled
#debug = false
# Disable color in terminal output
#disable_color = false
# Enable auto linting after editing
#enable_auto_lint = false
# Enable saving and restoring the session when run from CLI
#enable_cli_session = false
# File store path
#file_store_path = "/tmp/file_store"
# File store type
#file_store = "memory"
# List of allowed file extensions for uploads
#file_uploads_allowed_extensions = [".*"]
# Maximum file size for uploads, in megabytes
#file_uploads_max_file_size_mb = 0
# Maximum budget per task, 0.0 means no limit
#max_budget_per_task = 0.0
# Maximum number of iterations
#max_iterations = 100
# Path to mount the workspace in the sandbox
#workspace_mount_path_in_sandbox = "/workspace"
# Path to mount the workspace
#workspace_mount_path = ""
# Path to rewrite the workspace mount path to
#workspace_mount_rewrite = ""
# Persist the sandbox
persist_sandbox = false
# Run as devin
#run_as_devin = true
# Runtime environment
#runtime = "server"
# SSH hostname for the sandbox
#ssh_hostname = "localhost"
# SSH password for the sandbox
#ssh_password = ""
# SSH port for the sandbox
#ssh_port = 63710
# Use host network
#use_host_network = false
# Name of the default agent
#default_agent = "CodeActAgent"
#################################### LLM #####################################
# Configuration for LLM models (group name starts with 'llm')
# use 'llm' for the default LLM config
##############################################################################
[llm]
# AWS access key ID
#aws_access_key_id = ""
# AWS region name
#aws_region_name = ""
# AWS secret access key
#aws_secret_access_key = ""
# API key to use
api_key = "your-api-key"
# API base URL
#base_url = ""
# API version
#api_version = ""
# Cost per input token
#input_cost_per_token = 0.0
# Cost per output token
#output_cost_per_token = 0.0
# Custom LLM provider
#custom_llm_provider = ""
# Embedding API base URL
#embedding_base_url = ""
# Embedding deployment name
#embedding_deployment_name = ""
# Embedding model to use
embedding_model = ""
# Maximum number of characters in an observation's content
#max_message_chars = 10000
# Maximum number of input tokens
#max_input_tokens = 0
# Maximum number of output tokens
#max_output_tokens = 0
# Model to use
model = "gpt-4o"
# Number of retries to attempt
#num_retries = 5
# Retry maximum wait time
#retry_max_wait = 60
# Retry minimum wait time
#retry_min_wait = 3
# Temperature for the API
#temperature = 0.0
# Timeout for the API
#timeout = 0
# Top p for the API
#top_p = 0.5
[llm.gpt3]
# API key to use
api_key = "your-api-key"
# Model to use
model = "gpt-3.5"
#################################### Agent ###################################
# Configuration for agents (group name starts with 'agent')
# Use 'agent' for the default agent config
# otherwise, group name must be `agent.<agent_name>` (case-sensitive), e.g.
# agent.CodeActAgent
##############################################################################
[agent]
# Memory enabled
#memory_enabled = false
# Memory maximum threads
#memory_max_threads = 2
# LLM config group to use
#llm_config = 'llm'
[agent.RepoExplorerAgent]
# Example: use a cheaper model for RepoExplorerAgent to reduce cost, especially
# useful when an agent doesn't demand high quality but uses a lot of tokens
llm_config = 'gpt3'
#################################### Sandbox ###################################
# Configuration for the sandbox
##############################################################################
[sandbox]
# Sandbox timeout in seconds
#timeout = 120
# Sandbox type (ssh, e2b, local)
#box_type = "ssh"
# Sandbox user ID
#user_id = 1000
# Container image to use for the sandbox
#container_image = "ghcr.io/opendevin/sandbox:main"
#################################### Eval ####################################
# Configuration for the evaluation, please refer to the specific evaluation
# plugin for the available options
##############################################################################
+3 -2
View File
@@ -1,11 +1,12 @@
# Docker Containers
Each folder here contains a Dockerfile, and a config.sh describing how to build
the image and where to push it. These are images are built and pushed in GitHub Actions
the images and where to push them. These images are built and pushed in GitHub Actions
by the `ghcr.yml` workflow.
## Building Manually
```
```bash
docker build -f containers/app/Dockerfile -t opendevin .
docker build -f containers/sandbox/Dockerfile -t sandbox .
```
+32 -29
View File
@@ -3,65 +3,68 @@ set -eo pipefail
image_name=$1
org_name=$2
push=0
if [[ $3 == "--push" ]]; then
push=1
fi
platform=$3
echo -e "Building: $image_name"
echo "Building: $image_name for platform: $platform"
tags=()
OPEN_DEVIN_BUILD_VERSION="dev"
cache_tag_base="buildcache"
cache_tag="$cache_tag_base"
if [[ -n $GITHUB_REF_NAME ]]; then
# check if ref name is a version number
if [[ $GITHUB_REF_NAME =~ ^[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
major_version=$(echo $GITHUB_REF_NAME | cut -d. -f1)
minor_version=$(echo $GITHUB_REF_NAME | cut -d. -f1,2)
tags+=($major_version $minor_version)
major_version=$(echo "$GITHUB_REF_NAME" | cut -d. -f1)
minor_version=$(echo "$GITHUB_REF_NAME" | cut -d. -f1,2)
tags+=("$major_version" "$minor_version")
tags+=("latest")
fi
sanitized=$(echo $GITHUB_REF_NAME | sed 's/[^a-zA-Z0-9.-]\+/-/g')
sanitized=$(echo "$GITHUB_REF_NAME" | sed 's/[^a-zA-Z0-9.-]\+/-/g')
OPEN_DEVIN_BUILD_VERSION=$sanitized
cache_tag+="-${sanitized}"
tags+=($sanitized)
tag=$(echo "$sanitized" | tr '[:upper:]' '[:lower:]') # lower case is required in tagging
tags+=("$tag")
fi
echo "Tags: ${tags[@]}"
dir=./containers/$image_name
if [ ! -f $dir/Dockerfile ]; then
if [[ "$image_name" == "opendevin" ]]; then
dir="./containers/app"
else
dir="./containers/$image_name"
fi
if [[ ! -f "$dir/Dockerfile" ]]; then
echo "No Dockerfile found"
exit 1
fi
if [ ! -f $dir/config.sh ]; then
if [[ ! -f "$dir/config.sh" ]]; then
echo "No config.sh found for Dockerfile"
exit 1
fi
source $dir/config.sh
source "$dir/config.sh"
if [[ -n "$org_name" ]]; then
DOCKER_ORG="$org_name"
fi
DOCKER_REPOSITORY=$DOCKER_REGISTRY/$DOCKER_ORG/$DOCKER_IMAGE
DOCKER_REPOSITORY="$DOCKER_REGISTRY/$DOCKER_ORG/$DOCKER_IMAGE"
DOCKER_REPOSITORY=${DOCKER_REPOSITORY,,} # lowercase
echo "Repo: $DOCKER_REPOSITORY"
echo "Base dir: $DOCKER_BASE_DIR"
args=""
for tag in ${tags[@]}; do
for tag in "${tags[@]}"; do
args+=" -t $DOCKER_REPOSITORY:$tag"
done
if [[ $push -eq 1 ]]; then
args+=" --push"
args+=" --cache-to=type=registry,ref=$DOCKER_REPOSITORY:$cache_tag,mode=max"
fi
output_image="/tmp/${image_name}_image_${platform}.tar"
docker buildx build \
$args \
--build-arg OPEN_DEVIN_BUILD_VERSION=$OPEN_DEVIN_BUILD_VERSION \
--cache-from=type=registry,ref=$DOCKER_REPOSITORY:$cache_tag \
--cache-from=type=registry,ref=$DOCKER_REPOSITORY:$cache_tag_base-main \
--platform linux/amd64,linux/arm64 \
--build-arg OPEN_DEVIN_BUILD_VERSION="$OPEN_DEVIN_BUILD_VERSION" \
--platform linux/$platform \
--provenance=false \
-f $dir/Dockerfile $DOCKER_BASE_DIR
-f "$dir/Dockerfile" \
--output type=docker,dest="$output_image" \
"$DOCKER_BASE_DIR"
echo "${tags[*]}" > tags.txt
+14 -7
View File
@@ -16,13 +16,11 @@ RUN apt-get update && apt-get install -y \
build-essential \
openssh-server \
sudo \
bash \
gcc \
jq \
g++ \
make \
iproute2 \
libgl1-mesa-glx \
&& rm -rf /var/lib/apt/lists/*
RUN mkdir -p -m0755 /var/run/sshd
@@ -30,8 +28,17 @@ RUN mkdir -p -m0755 /var/run/sshd
# symlink python3 to python
RUN ln -s /usr/bin/python3 /usr/bin/python
# install basic dependencies for CodeActAgent
RUN pip3 install --upgrade pip
RUN pip3 install jupyterlab notebook jupyter_kernel_gateway flake8
# TODO: those dependencies are needed for agentskills, we should pack them in a new sandbox image
RUN pip3 install python-docx PyPDF2 python-pptx pylatexenc openai opencv-python
# ==== OpenDevin Runtime Client ====
RUN mkdir -p /opendevin && mkdir -p /opendevin/logs && chmod 777 /opendevin/logs
RUN wget --progress=bar:force -O Miniforge3.sh "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
RUN bash Miniforge3.sh -b -p /opendevin/miniforge3
RUN chmod -R g+w /opendevin/miniforge3
RUN bash -c ". /opendevin/miniforge3/etc/profile.d/conda.sh && conda config --set changeps1 False && conda config --append channels conda-forge"
RUN echo "" > /opendevin/bash.bashrc
RUN rm -f Miniforge3.sh
# - agentskills dependencies
RUN /opendevin/miniforge3/bin/pip install --upgrade pip
RUN /opendevin/miniforge3/bin/pip install jupyterlab notebook jupyter_kernel_gateway flake8
RUN /opendevin/miniforge3/bin/pip install python-docx PyPDF2 python-pptx pylatexenc openai
RUN /opendevin/miniforge3/bin/pip install python-dotenv toml termcolor pydantic python-docx pyyaml docker pexpect tenacity e2b browsergym minio
+13 -4
View File
@@ -8,8 +8,8 @@ const config: Config = {
favicon: "img/logo.png",
// Set the production url of your site here
url: "https://OpenDevin.github.io",
baseUrl: "/OpenDevin/",
url: "https://docs.all-hands.dev",
baseUrl: "/",
// GitHub pages deployment config.
organizationName: "OpenDevin",
@@ -23,8 +23,13 @@ const config: Config = {
// useful metadata like html lang. For example, if your site is Chinese, you
// may want to replace "en" with "zh-Hans".
i18n: {
defaultLocale: "en",
locales: ["en"],
defaultLocale: 'en',
locales: ['en', 'fr', 'zh-Hans'],
localeConfigs: {
en: {
htmlLang: 'en-GB',
},
},
},
presets: [
@@ -78,6 +83,10 @@ const config: Config = {
label: "GitHub",
position: "right",
},
{
type: 'localeDropdown',
position: 'left',
},
],
},
prism: {
+406
View File
@@ -0,0 +1,406 @@
{
"footer.title": {
"message": "OpenDevin"
},
"footer.docs": {
"message": "Documents"
},
"footer.community": {
"message": "Communauté"
},
"footer.copyright": {
"message": "© {year} OpenDevin"
},
"faq.title": {
"message": "Questions Fréquemment Posées",
"description": "FAQ Title"
},
"faq.description": {
"message": "Questions Fréquemment Posées"
},
"faq.section.title.1": {
"message": "Qu'est-ce qu'OpenDevin ?",
"description": "First Section Title"
},
"faq.section.highlight": {
"message": "OpenDevin",
"description": "Highlight Text"
},
"faq.section.description.1": {
"message": "est un ingénieur logiciel autonome qui peut résoudre des tâches d'ingénierie logicielle et de navigation web à tout moment. Il peut exécuter des requêtes en sciences des données, telles que \"Trouver le nombre de demandes de pull à l'repository OpenDevin dans les derniers mois\", et des tâches d'ingénierie logicielle, comme \"Veuillez ajouter des tests à ce fichier et vérifier si tous les tests passent. Si ce n'est pas le cas, réparez le fichier.\"",
"description": "Description for OpenDevin"
},
"faq.section.description.2": {
"message": "De plus, OpenDevin est une plateforme et communauté pour les développeurs d'agents qui souhaitent tester et évaluer de nouveaux agents.",
"description": "Further Description for OpenDevin"
},
"faq.section.title.2": {
"message": "Support",
"description": "Support Section Title"
},
"faq.section.support.answer": {
"message": "Si vous rencontrez un problème que d'autres utilisateurs peuvent également avoir, merci de le signaler sur {githubLink}. Si vous avez des difficultés à l'installation ou des questions générales, rejoignez-vous sur {discordLink} ou {slackLink}.",
"description": "Support Answer"
},
"faq.section.title.3": {
"message": "Comment résoudre un problème sur GitHub avec OpenDevin ?",
"description": "GitHub Issue Section Title"
},
"faq.section.github.steps.intro": {
"message": "Pour résoudre un problème sur GitHub en utilisant OpenDevin, envoyez une commande à OpenDevin demandant qu'il suit des étapes comme les suivantes :",
"description": "GitHub Steps Introduction"
},
"faq.section.github.step1": {
"message": "Lisez l'issue https://github.com/OpenDevin/OpenDevin/issues/1611",
"description": "GitHub Step 1"
},
"faq.section.github.step2": {
"message": "Cloner le dépôt et vérifier une nouvelle branche",
"description": "GitHub Step 2"
},
"faq.section.github.step3": {
"message": "Sur la base des instructions dans la description de l'issue, modifiez les fichiers pour résoudre le problème",
"description": "GitHub Step 3"
},
"faq.section.github.step4": {
"message": "Pousser le résultat à GitHub en utilisant la variable d'environnement GITHUB_TOKEN",
"description": "GitHub Step 4"
},
"faq.section.github.step5": {
"message": "Dites-moi le lien que je dois utiliser pour envoyer une demande de pull",
"description": "GitHub Step 5"
},
"faq.section.github.steps.preRun": {
"message": "Avant de lancer OpenDevin, vous pouvez faire :",
"description": "GitHub Steps Pre-Run"
},
"faq.section.github.steps.tokenInfo": {
"message": "où XXX est un jeton GitHub que vous avez créé et qui a les autorisations pour pousser dans le dépôt OpenDevin. Si vous n'avez pas d'autorisations de modification du dépôt OpenDevin, vous devrez peut-être changer cela en :",
"description": "GitHub Steps Token Info"
},
"faq.section.github.steps.usernameInfo": {
"message": "où USERNAME est votre nom GitHub.",
"description": "GitHub Steps Username Info"
},
"faq.section.title.4": {
"message": "Comment OpenDevin est-il différent de Devin ?",
"description": "Devin Section Title"
},
"faq.section.devin.linkText": {
"message": "Devin",
"description": "Devin Link Text"
},
"faq.section.devin.description": {
"message": "est un produit commercial par Cognition Inc., qui a servi d'inspiration initiale pour OpenDevin. Les deux visent à bien faire le travail d'ingénierie logicielle, mais vous pouvez télécharger, utiliser et modifier OpenDevin, tandis que Devin peut être utilisé uniquement via le site de Cognition. De plus, OpenDevin a évolué au-delà de l'inspiration initiale, et est maintenant un écosystème communautaire pour le développement d'agents en général, et nous serions ravis de vous voir rejoindre et",
"description": "Devin Description"
},
"faq.section.devin.contribute": {
"message": "contribuer",
"description": "Contribute Link"
},
"faq.section.title.5": {
"message": "Comment OpenDevin est-il différent de ChatGPT ?",
"description": "ChatGPT Section Title"
},
"faq.section.chatgpt.description": {
"message": "ChatGPT vous pouvez accéder en ligne, il ne se connecte pas aux fichiers locaux et ses capacités d'exécution du code sont limitées. Alors qu'il peut écrire du code, mais c'est difficile à tester ou à exécuter.",
"description": "ChatGPT Description"
},
"homepage.description": {
"message": "Génération d'code AI pour l'ingénierie logicielle.",
"description": "The homepage description"
},
"homepage.getStarted": {
"message": "Commencer"
},
"welcome.message": {
"message": "Bienvenue à OpenDevin, un système d'IA autonome ingénieur logiciel capable d'exécuter des tâches d'ingénierie complexes et de collaborer activement avec les utilisateurs sur les projets de développement logiciel."
},
"theme.ErrorPageContent.title": {
"message": "Cette page a planté.",
"description": "The title of the fallback page when the page crashed"
},
"theme.BackToTopButton.buttonAriaLabel": {
"message": "Retourner en haut de la page",
"description": "The ARIA label for the back to top button"
},
"theme.blog.archive.title": {
"message": "Archives",
"description": "The page & hero title of the blog archive page"
},
"theme.blog.archive.description": {
"message": "Archives",
"description": "The page & hero description of the blog archive page"
},
"theme.blog.paginator.navAriaLabel": {
"message": "Pagination des listes d'articles du blog",
"description": "The ARIA label for the blog pagination"
},
"theme.blog.paginator.newerEntries": {
"message": "Nouvelles entrées",
"description": "The label used to navigate to the newer blog posts page (previous page)"
},
"theme.blog.paginator.olderEntries": {
"message": "Anciennes entrées",
"description": "The label used to navigate to the older blog posts page (next page)"
},
"theme.blog.post.paginator.navAriaLabel": {
"message": "Pagination des articles du blog",
"description": "The ARIA label for the blog posts pagination"
},
"theme.blog.post.paginator.newerPost": {
"message": "Article plus récent",
"description": "The blog post button label to navigate to the newer/previous post"
},
"theme.blog.post.paginator.olderPost": {
"message": "Article plus ancien",
"description": "The blog post button label to navigate to the older/next post"
},
"theme.blog.post.plurals": {
"message": "Un article|{count} articles",
"description": "Pluralized label for \"{count} posts\". Use as much plural forms (separated by \"|\") as your language support (see https://www.unicode.org/cldr/cldr-aux/charts/34/supplemental/language_plural_rules.html)"
},
"theme.blog.tagTitle": {
"message": "{nPosts} tags avec « {tagName} »",
"description": "The title of the page for a blog tag"
},
"theme.tags.tagsPageLink": {
"message": "Voir tous les tags",
"description": "The label of the link targeting the tag list page"
},
"theme.colorToggle.ariaLabel": {
"message": "Basculer entre le mode sombre et clair (actuellement {mode})",
"description": "The ARIA label for the navbar color mode toggle"
},
"theme.colorToggle.ariaLabel.mode.dark": {
"message": "mode sombre",
"description": "The name for the dark color mode"
},
"theme.colorToggle.ariaLabel.mode.light": {
"message": "mode clair",
"description": "The name for the light color mode"
},
"theme.docs.breadcrumbs.navAriaLabel": {
"message": "Bouton de navigation des liens de la page",
"description": "The ARIA label for the breadcrumbs"
},
"theme.docs.DocCard.categoryDescription.plurals": {
"message": "1 élément|{count} éléments",
"description": "The default description for a category card in the generated index about how many items this category includes"
},
"theme.docs.paginator.navAriaLabel": {
"message": "Pages de documentation",
"description": "The ARIA label for the docs pagination"
},
"theme.docs.paginator.previous": {
"message": "Précédent",
"description": "The label used to navigate to the previous doc"
},
"theme.docs.paginator.next": {
"message": "Suivant",
"description": "The label used to navigate to the next doc"
},
"theme.docs.tagDocListPageTitle.nDocsTagged": {
"message": "Un document tagué|{count} documents tagués",
"description": "Pluralized label for \"{count} docs tagged\". Use as much plural forms (separated by \"|\") as your language support (see https://www.unicode.org/cldr/cldr-aux/charts/34/supplemental/language_plural_rules.html)"
},
"theme.docs.tagDocListPageTitle": {
"message": "{nDocsTagged} avec \"{tagName}\"",
"description": "The title of the page for a docs tag"
},
"theme.docs.versionBadge.label": {
"message": "Version: {versionLabel}"
},
"theme.docs.versions.unreleasedVersionLabel": {
"message": "Ceci est la documentation de la prochaine version {versionLabel} de {siteTitle}.",
"description": "The label used to tell the user that he's browsing an unreleased doc version"
},
"theme.docs.versions.unmaintainedVersionLabel": {
"message": "Ceci est la documentation de {siteTitle} {versionLabel}, qui n'est plus activement maintenue.",
"description": "The label used to tell the user that he's browsing an unmaintained doc version"
},
"theme.docs.versions.latestVersionSuggestionLabel": {
"message": "Pour une documentation à jour, consultez la {latestVersionLink} ({versionLabel}).",
"description": "The label used to tell the user to check the latest version"
},
"theme.docs.versions.latestVersionLinkLabel": {
"message": "dernière version",
"description": "The label used for the latest version suggestion link label"
},
"theme.common.editThisPage": {
"message": "Éditer cette page",
"description": "The link label to edit the current page"
},
"theme.common.headingLinkTitle": {
"message": "Lien direct vers {heading}",
"description": "Title for link to heading"
},
"theme.lastUpdated.atDate": {
"message": " le {date}",
"description": "The words used to describe on which date a page has been last updated"
},
"theme.lastUpdated.byUser": {
"message": " par {user}",
"description": "The words used to describe by who the page has been last updated"
},
"theme.lastUpdated.lastUpdatedAtBy": {
"message": "Dernière mise à jour{atDate}{byUser}",
"description": "The sentence used to display when a page has been last updated, and by who"
},
"theme.navbar.mobileVersionsDropdown.label": {
"message": "Versions",
"description": "The label for the navbar versions dropdown on mobile view"
},
"theme.NotFound.title": {
"message": "Page introuvable",
"description": "The title of the 404 page"
},
"theme.tags.tagsListLabel": {
"message": "Tags :",
"description": "The label alongside a tag list"
},
"theme.admonition.caution": {
"message": "prudence",
"description": "The default label used for the Caution admonition (:::caution)"
},
"theme.admonition.danger": {
"message": "danger",
"description": "The default label used for the Danger admonition (:::danger)"
},
"theme.admonition.info": {
"message": "information",
"description": "The default label used for the Info admonition (:::info)"
},
"theme.admonition.note": {
"message": "remarque",
"description": "The default label used for the Note admonition (:::note)"
},
"theme.admonition.tip": {
"message": "astuce",
"description": "The default label used for the Tip admonition (:::tip)"
},
"theme.admonition.warning": {
"message": "prudence",
"description": "The default label used for the Warning admonition (:::warning)"
},
"theme.AnnouncementBar.closeButtonAriaLabel": {
"message": "Fermer",
"description": "The ARIA label for close button of announcement bar"
},
"theme.blog.sidebar.navAriaLabel": {
"message": "Navigation vers les articles récents du blog",
"description": "The ARIA label for recent posts in the blog sidebar"
},
"theme.CodeBlock.copied": {
"message": "Copié",
"description": "The copied button label on code blocks"
},
"theme.CodeBlock.copyButtonAriaLabel": {
"message": "Copier le code",
"description": "The ARIA label for copy code blocks button"
},
"theme.CodeBlock.copy": {
"message": "Copier",
"description": "The copy button label on code blocks"
},
"theme.CodeBlock.wordWrapToggle": {
"message": "Activer/désactiver le retour à la ligne",
"description": "The title attribute for toggle word wrapping button of code block lines"
},
"theme.DocSidebarItem.expandCategoryAriaLabel": {
"message": "Développer la catégorie '{label}' de la barre latérale",
"description": "The ARIA label to expand the sidebar category"
},
"theme.DocSidebarItem.collapseCategoryAriaLabel": {
"message": "Réduire la catégorie '{label}' de la barre latérale",
"description": "The ARIA label to collapse the sidebar category"
},
"theme.NavBar.navAriaLabel": {
"message": "Main",
"description": "The ARIA label for the main navigation"
},
"theme.navbar.mobileLanguageDropdown.label": {
"message": "Langues",
"description": "The label for the mobile language switcher dropdown"
},
"theme.NotFound.p1": {
"message": "Nous n'avons pas trouvé ce que vous recherchez.",
"description": "The first paragraph of the 404 page"
},
"theme.NotFound.p2": {
"message": "Veuillez contacter le propriétaire du site qui vous a lié à l'URL d'origine et leur faire savoir que leur lien est cassé.",
"description": "The 2nd paragraph of the 404 page"
},
"theme.TOCCollapsible.toggleButtonLabel": {
"message": "Sur cette page",
"description": "The label used by the button on the collapsible TOC component"
},
"theme.blog.post.readMore": {
"message": "Lire plus",
"description": "The label used in blog post item excerpts to link to full blog posts"
},
"theme.blog.post.readMoreLabel": {
"message": "En savoir plus sur {title}",
"description": "The ARIA label for the link to full blog posts from excerpts"
},
"theme.blog.post.readingTime.plurals": {
"message": "Une minute de lecture|{readingTime} minutes de lecture",
"description": "Pluralized label for \"{readingTime} min read\". Use as much plural forms (separated by \"|\") as your language support (see https://www.unicode.org/cldr/cldr-aux/charts/34/supplemental/language_plural_rules.html)"
},
"theme.docs.breadcrumbs.home": {
"message": "Page d'accueil",
"description": "The ARIA label for the home page in the breadcrumbs"
},
"theme.docs.sidebar.collapseButtonTitle": {
"message": "Réduire le menu latéral",
"description": "The title attribute for collapse button of doc sidebar"
},
"theme.docs.sidebar.collapseButtonAriaLabel": {
"message": "Réduire le menu latérale",
"description": "The title attribute for collapse button of doc sidebar"
},
"theme.docs.sidebar.navAriaLabel": {
"message": "Barre de navigation latérale des docs",
"description": "The ARIA label for the sidebar navigation"
},
"theme.docs.sidebar.closeSidebarButtonAriaLabel": {
"message": "Fermer la barre de navigation",
"description": "The ARIA label for close button of mobile sidebar"
},
"theme.navbar.mobileSidebarSecondaryMenu.backButtonLabel": {
"message": "← Retour au menu principal",
"description": "The label of the back button to return to main menu, inside the mobile navbar sidebar secondary menu (notably used to display the docs sidebar)"
},
"theme.docs.sidebar.toggleSidebarButtonAriaLabel": {
"message": "Ouvrir/fermer la barre de navigation",
"description": "The ARIA label for hamburger menu button of mobile navigation"
},
"theme.docs.sidebar.expandButtonTitle": {
"message": "Déplier le menu latéral",
"description": "The ARIA label and title attribute for expand button of doc sidebar"
},
"theme.docs.sidebar.expandButtonAriaLabel": {
"message": "Déployer le menu latérale",
"description": "The ARIA label and title attribute for expand button of doc sidebar"
},
"theme.ErrorPageContent.tryAgain": {
"message": "Réessayer",
"description": "The label of the button to try again rendering when the React error boundary captures an error"
},
"theme.common.skipToMainContent": {
"message": "Aller directement au contenu principal",
"description": "The skip to content label used for accessibility, allowing to rapidly navigate to main content with keyboard tab/enter navigation"
},
"theme.tags.tagsPageTitle": {
"message": "Tags",
"description": "The title of the tag list page"
},
"theme.unlistedContent.title": {
"message": "Page non répertoriée",
"description": "The unlisted content banner title"
},
"theme.unlistedContent.message": {
"message": "Cette page n'est pas répertoriée. Les moteurs de recherche ne l'indexeront pas, et seuls les utilisateurs ayant un lien direct peuvent y accéder.",
"description": "The unlisted content banner message"
}
}
@@ -0,0 +1,14 @@
{
"title": {
"message": "Blog",
"description": "The title for the blog used in SEO"
},
"description": {
"message": "Blog",
"description": "The description for the blog used in SEO"
},
"sidebar.title": {
"message": "Articles récents",
"description": "The label for the left sidebar"
}
}
@@ -0,0 +1,18 @@
{
"version.label": {
"message": "Next",
"description": "The label for version current"
},
"sidebar.docsSidebar.category.🤖 Backends LLM": {
"message": "🤖 Backends LLM",
"description": "The label for category 🤖 Backends LLM in sidebar docsSidebar"
},
"sidebar.docsSidebar.category.🚧 Dépannage": {
"message": "🚧 Dépannage",
"description": "The label for category 🚧 Dépannage in sidebar docsSidebar"
},
"sidebar.apiSidebar.category.Backend": {
"message": "Backend",
"description": "The label for category Backend in sidebar apiSidebar"
}
}
@@ -0,0 +1,3 @@
# Documentation Python
Les documents apparaîtront ici après le déploiement.
@@ -0,0 +1,5 @@
{
"items": ["python/python"],
"label": "Backend",
"type": "categorie"
}
@@ -0,0 +1,53 @@
---
sidebar_position: 7
---
# 📚 Divers {#misc}
## ⭐️ Stratégie de Recherche {#research-strategy}
La réalisation d'une réplication complète des applications de production avec les LLM est une entreprise complexe. Notre stratégie implique :
1. **Recherche Technique de Base :** Se concentrer sur la recherche fondamentale pour comprendre et améliorer les aspects techniques de la génération et de la gestion de code.
2. **Compétences Spécialisées :** Améliorer l'efficacité des composants de base grâce à la curation des données, aux méthodes de formation, et plus encore.
3. **Planification des Tâches :** Développer des capacités pour la détection de bogues, la gestion du code source et l'optimisation.
4. **Évaluation :** Établir des métriques d'évaluation complètes pour mieux comprendre et améliorer nos modèles.
## 🚧 Agent Par Défaut {#default-agent}
- Notre agent par défaut est actuellement le CodeActAgent, capable de générer du code et de gérer des fichiers. Nous travaillons sur d'autres implémentations d'agents, y compris [SWE Agent](https://swe-agent.com/). Vous pouvez [lire à propos de notre ensemble actuel d'agents ici](./agents).
## 🤝 Comment Contribuer {#how-to-contribute}
OpenDevin est un projet communautaire, et nous accueillons les contributions de tout le monde. Que vous soyez développeur, chercheur, ou simplement enthousiaste à l'idée de faire progresser le domaine de l'ingénierie logicielle avec l'IA, il existe de nombreuses façons de vous impliquer :
- **Contributions de Code :** Aidez-nous à développer les fonctionnalités de base, l'interface frontend ou les solutions de sandboxing.
- **Recherche et Évaluation :** Contribuez à notre compréhension des LLM en ingénierie logicielle, participez à l'évaluation des modèles ou suggérez des améliorations.
- **Retour d'Information et Tests :** Utilisez l'ensemble d'outils OpenDevin, signalez des bogues, suggérez des fonctionnalités ou fournissez des retours sur l'ergonomie.
Pour plus de détails, veuillez consulter [ce document](https://github.com/OpenDevin/OpenDevin/blob/main/CONTRIBUTING.md).
## 🤖 Rejoignez Notre Communauté {#join-our-community}
Nous avons maintenant à la fois un espace de travail Slack pour la collaboration sur la construction d'OpenDevin et un serveur Discord pour discuter de tout ce qui est lié, par exemple, à ce projet, aux LLM, aux agents, etc.
- [Espace de travail Slack](https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA)
- [Serveur Discord](https://discord.gg/ESHStjSjD4)
Si vous souhaitez contribuer, n'hésitez pas à rejoindre notre communauté. Simplifions l'ingénierie logicielle ensemble !
🐚 **Codez moins, créez plus avec OpenDevin.**
[![Star History Chart](https://api.star-history.com/svg?repos=OpenDevin/OpenDevin&type=Date)](https://star-history.com/#OpenDevin/OpenDevin&Date)
## 🛠️ Construit Avec {#built-with}
OpenDevin est construit en utilisant une combinaison de cadres et de bibliothèques puissants, offrant une base robuste pour son développement. Voici les technologies clés utilisées dans le projet :
![FastAPI](https://img.shields.io/badge/FastAPI-black?style=for-the-badge) ![uvicorn](https://img.shields.io/badge/uvicorn-black?style=for-the-badge) ![LiteLLM](https://img.shields.io/badge/LiteLLM-black?style=for-the-badge) ![Docker](https://img.shields.io/badge/Docker-black?style=for-the-badge) ![Ruff](https://img.shields.io/badge/Ruff-black?style=for-the-badge) ![MyPy](https://img.shields.io/badge/MyPy-black?style=for-the-badge) ![LlamaIndex](https://img.shields.io/badge/LlamaIndex-black?style=for-the-badge) ![React](https://img.shields.io/badge/React-black?style=for-the-badge)
Veuillez noter que la sélection de ces technologies est en cours, et que des technologies supplémentaires peuvent être ajoutées ou des existantes supprimées au fur et à mesure de l'évolution du projet. Nous nous efforçons d'adopter les outils les plus adaptés et efficaces pour améliorer les capacités d'OpenDevin.
## 📜 Licence {#license}
Distribué sous la licence MIT. Voir [notre licence](https://github.com/OpenDevin/OpenDevin/blob/main/LICENSE) pour plus d'informations.
@@ -0,0 +1,134 @@
---
sidebar_position: 3
---
# 🧠 Agents et Capacités
## Agent CodeAct
### Description
Cet agent implémente l'idée CodeAct ([article](https://arxiv.org/abs/2402.01030), [tweet](https://twitter.com/xingyaow_/status/1754556835703751087)) qui consolide les **act**ions des agents LLM en un espace d'action **code** unifié pour à la fois la _simplicité_ et la _performance_ (voir article pour plus de détails).
L'idée conceptuelle est illustrée ci-dessous. À chaque tour, l'agent peut :
1. **Converse** : Communiquer avec les humains en langage naturel pour demander des clarifications, des confirmations, etc.
2. **CodeAct** : Choisir d'accomplir la tâche en exécutant du code
- Exécuter toute commande `bash` Linux valide
- Exécuter tout code `Python` valide avec [un interpréteur Python interactif](https://ipython.org/). Cela est simulé à travers la commande `bash`, voir le système de plugin ci-dessous pour plus de détails.
![image](https://github.com/OpenDevin/OpenDevin/assets/38853559/92b622e3-72ad-4a61-8f41-8c040b6d5fb3)
### Système de Plugin
Pour rendre l'agent CodeAct plus puissant avec seulement l'accès à l'espace d'action `bash`, l'agent CodeAct exploite le système de plugins d'OpenDevin:
- [Plugin Jupyter](https://github.com/OpenDevin/OpenDevin/tree/main/opendevin/runtime/plugins/jupyter) : pour l'exécution d'IPython via la commande bash
- [Plugin outil agent SWE](https://github.com/OpenDevin/OpenDevin/tree/main/opendevin/runtime/plugins/swe_agent_commands) : Outils de ligne de commande bash puissants pour les tâches de développement logiciel introduits par [swe-agent](https://github.com/princeton-nlp/swe-agent).
### Démonstration
https://github.com/OpenDevin/OpenDevin/assets/38853559/f592a192-e86c-4f48-ad31-d69282d5f6ac
_Exemple de CodeActAgent avec `gpt-4-turbo-2024-04-09` effectuant une tâche de science des données (régression linéaire)_
### Actions
`Action`,
`CmdRunAction`,
`IPythonRunCellAction`,
`AgentEchoAction`,
`AgentFinishAction`,
`AgentTalkAction`
### Observations
`CmdOutputObservation`,
`IPythonRunCellObservation`,
`AgentMessageObservation`,
`UserMessageObservation`
### Méthodes
| Méthode | Description |
| ---------------- | ----------------------------------------------------------------------------------------------------------------------------------------------- |
| `__init__` | Initialise un agent avec `llm` et une liste de messages `list[Mapping[str, str]]` |
| `step` | Effectue une étape en utilisant l'agent CodeAct. Cela inclut la collecte d'informations sur les étapes précédentes et invite le modèle à exécuter une commande. |
### En cours de réalisation & prochaine étape
[] Support de la navigation sur le web
[] Compléter le workflow pour l'agent CodeAct afin de soumettre des PRs Github
## Agent Monologue
### Description
L'agent Monologue utilise la mémoire à long terme et à court terme pour accomplir des tâches.
La mémoire à long terme est stockée en tant qu'objet LongTermMemory et le modèle l'utilise pour rechercher des exemples du passé.
La mémoire à court terme est stockée en tant qu'objet Monologue et le modèle peut la condenser si nécessaire.
### Actions
`Action`,
`NullAction`,
`CmdRunAction`,
`FileWriteAction`,
`FileReadAction`,
`BrowseURLAction`,
`GithubPushAction`,
`AgentThinkAction`
### Observations
`Observation`,
`NullObservation`,
`CmdOutputObservation`,
`FileReadObservation`,
`BrowserOutputObservation`
### Méthodes
| Méthode | Description |
| --------------- | --------------------------------------------------------------------------------------------------------------------------------------------- |
| `__init__` | Initialise l'agent avec une mémoire à long terme et un monologue interne |
| `_add_event` | Ajoute des événements au monologue de l'agent et condense avec un résumé automatiquement si le monologue est trop long |
| `_initialize` | Utilise la liste `INITIAL_THOUGHTS` pour donner à l'agent un contexte pour ses capacités et comment naviguer dans le `/workspace` |
| `step` | Modifie l'état actuel en ajoutant les actions et observations les plus récentes, puis invite le modèle à réfléchir à la prochaine action à entreprendre. |
## Agent Planificateur
### Description
L'agent planificateur utilise une stratégie d'incitation spéciale pour créer des plans à long terme pour résoudre les problèmes.
L'agent reçoit ses paires action-observation précédentes, la tâche actuelle, et un indice basé sur la dernière action effectuée à chaque étape.
### Actions
`NullAction`,
`CmdRunAction`,
`BrowseURLAction`,
`GithubPushAction`,
`FileReadAction`,
`FileWriteAction`,
`AgentThinkAction`,
`AgentFinishAction`,
`AgentSummarizeAction`,
`AddTaskAction`,
`ModifyTaskAction`,
### Observations
`Observation`,
`NullObservation`,
`CmdOutputObservation`,
`FileReadObservation`,
`BrowserOutputObservation`
### Méthodes
| Méthode | Description |
| ---------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `__init__` | Initialise un agent avec `llm` |
| `step` | Vérifie si l'étape actuelle est terminée, retourne `AgentFinishAction` si oui. Sinon, crée une incitation de planification et l'envoie au modèle pour inférence, en ajoutant le résultat comme prochaine action. |
@@ -0,0 +1,50 @@
---
sidebar_position: 4
---
# 🏛️ Aperçu de l'Architecture Système
Voici un aperçu de haut niveau de l'architecture du système. Le système est divisé en deux composants principaux : le frontend et le backend. Le frontend est responsable de la gestion des interactions avec l'utilisateur et de l'affichage des résultats. Le backend est responsable de la gestion de la logique métier et de l'exécution des agents.
![system_architecture.svg](/img/system_architecture.svg)
Cet aperçu est simplifié pour montrer les principaux composants et leurs interactions. Pour une vue plus détaillée de l'architecture du backend, consultez la section [Architecture du Backend](#backend-architecture-fr).
# Architecture du Backend {#backend-architecture-fr}
_**Avertissement**: L'architecture du backend est en cours de développement et est sujette à modifications. Le schéma suivant montre l'architecture actuelle du backend basée sur le commit indiqué dans le pied de page du schéma._
![backend_architecture.svg](/img/backend_architecture.svg)
<details>
<summary>Mise à jour de ce Schéma</summary>
<div>
La génération du schéma d'architecture du backend est partiellement automatisée.
Le schéma est généré à partir des annotations de type dans le code en utilisant l'outil py2puml.
Le schéma est ensuite revu manuellement, ajusté et exporté en PNG et SVG.
## Prérequis
- Un environnement Python dans lequel opendevin est exécutable
(selon les instructions du fichier README.md à la racine du dépôt)
- [py2puml](https://github.com/lucsorel/py2puml) installé
## Étapes
1. Générez automatiquement le schéma en exécutant la commande suivante depuis la racine du dépôt :
`py2puml opendevin opendevin > docs/architecture/backend_architecture.puml`
2. Ouvrez le fichier généré dans un éditeur PlantUML, par exemple Visual Studio Code avec l'extension PlantUML ou [PlantText](https://www.planttext.com/)
3. Révisez le PUML généré et apportez toutes les modifications nécessaires au schéma (ajoutez les parties manquantes, corrigez les erreurs, améliorez l'agencement).
_py2puml crée le schéma à partir des annotations de type dans le code, donc les annotations de type manquantes ou incorrectes peuvent entraîner un schéma incomplet ou incorrect._
4. Examinez la différence entre le nouveau schéma et le précédent et vérifiez manuellement si les modifications sont correctes.
_Assurez-vous de ne pas supprimer les parties ajoutées manuellement au schéma par le passé et qui sont toujours pertinentes._
5. Ajoutez le hash du commit qui a été utilisé pour générer le schéma dans le pied de page du schéma.
6. Exporte le schéma sous forme de fichiers PNG et SVG et remplacez les schémas existants dans le répertoire `docs/architecture`. Cela peut être fait avec (par exemple [PlantText](https://www.planttext.com/))
</div>
</details>
@@ -0,0 +1,107 @@
# 💿 Comment Créer un Soutien Docker sur Mesure
Le sandbox par défaut OpenDevin est équipé d'une configuration ubuntu minimaliste. Votre cas d'utilisation pourrait nécessiter des logiciels installés par défaut. Cet article vous enseignera comment réaliser cela en utilisant une image docker personnalisée.
## Configuration
Assurez-vous de pouvoir utiliser OpenDevin en suivant la documentation [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md).
## Créer Votre Image Docker
Ensuite, vous devez créer votre image docker personnalisée qui doit être basée sur debian/ubuntu. Par exemple, si nous souhaitons que OpenDevin ait accès au "node" binaire, nous utiliserions ce Dockerfile:
```bash
# Commencez avec l'image ubuntu la plus récente
FROM ubuntu:latest
# Effectuez les mises à jour nécessaires
RUN apt-get update && apt-get install
# Installez nodejs
RUN apt-get install -y nodejs
```
Ensuite, construisez votre image docker avec le nom de votre choix. Par exemple "image_personnalisée". Pour cela, créez un répertoire et placez le fichier à l'intérieur avec le nom "Dockerfile", puis dans le répertoire exécutez cette commande:
```bash
docker build -t image_personnalisée .
```
Cela produira une nouvelle image appelée ```image_personnalisée``` qui sera disponible dans Docker Engine.
> Remarque: Dans la configuration décrite ici, OpenDevin va fonctionner en tant que utilisateur "opendevin" à l'intérieur du sandbox et donc tous les packages installés via le Dockerfile seront disponibles pour tous les utilisateurs sur le système, pas seulement root.
>
> L'installation avec apt-get ci-dessus installe nodejs pour tous les utilisateurs.
## Spécifiez votre image personnalisée dans le fichier config.toml
La configuration OpenDevin se fait via le fichier de niveau supérieur ```config.toml``` .
Créez un fichier ```config.toml``` dans le répertoire OpenDevin et entrez ces contenus:
```toml
[core]
workspace_base="./workspace"
persist_sandbox=false
run_as_devin=true
sandbox_container_image="image_personnalisée"
```
> Assurez-vous que ```sandbox_container_image``` est défini sur le nom de votre image personnalisée précédente.
## Exécution
Exécutez OpenDevin en exécutant ```make run``` dans le répertoire racine.
Naviguez vers ```localhost:3001``` et vérifiez si vos dépendances souhaitées sont disponibles.
Dans le cas de l'exemple ci-dessus, la commande ```node -v``` dans la console produit ```v18.19.1```
Félicitations !
## Explication technique
Le code pertinent est défini dans [ssh_box.py](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/ssh_box.py) et [image_agnostic_util.py](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py).
En particulier, ssh_box.py vérifie l'objet config pour ```config.sandbox_container_image``` et ensuite tente de récupérer l'image à l'aide de [get_od_sandbox_image](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L72), qui est défini dans image_agnostic_util.py.
Lorsqu'une image personnalisée est utilisée pour la première fois, elle ne sera pas trouvée et donc elle sera construite (à l'exécution ultérieure, l'image construite sera trouvée et renvoyée).
L'image personnalisée est construite avec [_build_sandbox_image()](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L29), qui crée un fichier docker en utilisant votre image personnalisée comme base et configure ensuite l'environnement pour OpenDevin, comme ceci:
```python
dockerfile_content = (
f'FROM {base_image}\n'
'RUN apt update && apt install -y openssh-server wget sudo\n'
'RUN mkdir -p -m0755 /var/run/sshd\n'
'RUN mkdir -p /opendevin && mkdir -p /opendevin/logs && chmod 777 /opendevin/logs\n'
'RUN wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"\n'
'RUN bash Miniforge3-$(uname)-$(uname -m).sh -b -p /opendevin/miniforge3\n'
'RUN bash -c ". /opendevin/miniforge3/etc/profile.d/conda.sh && conda config --set changeps1 False && conda config --append channels conda-forge"\n'
'RUN echo "export PATH=/opendevin/miniforge3/bin:$PATH" >> ~/.bashrc\n'
'RUN echo "export PATH=/opendevin/miniforge3/bin:$PATH" >> /opendevin/bash.bashrc\n'
).strip()
```
> Remarque: Le nom de l'image est modifié via [_get_new_image_name()](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L63) et c'est ce nom modifié qui sera recherché lors des exécutions ultérieures.
## Dépannage / Erreurs
### Erreur: ```useradd: UID 1000 est non unique```
Si vous voyez cette erreur dans la sortie de la console, il s'agit du fait que OpenDevin essaie de créer le utilisateur opendevin dans le sandbox avec un ID d'utilisateur de 1000, cependant cet ID d'utilisateur est déjà utilisé dans l'image (pour une raison inconnue). Pour résoudre ce problème, changez la valeur du champ sandbox_user_id dans le fichier config.toml en une valeur différente:
```toml
[core]
workspace_base="./workspace"
persist_sandbox=false
run_as_devin=true
sandbox_container_image="image_personnalisée"
sandbox_user_id="1001"
```
### Erreurs de port d'utilisation
Si vous voyez un message d'erreur indiquant que le port est utilisé ou indisponible, essayez de supprimer toutes les containers docker en cours d'exécution (exécutez `docker ps` et `docker rm` des containers concernés) puis ré-exécutez ```make run```
## Discuter
Pour d'autres problèmes ou questions rejoignez le [Slack](https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA) ou le [Discord](https://discord.gg/ESHStjSjD4) et demandez!
@@ -0,0 +1,18 @@
---
sidebar_position: 6
---
# ✅ Fournir des Commentaires
Lorsque vous utilisez OpenDevin, vous rencontrerez sans aucun doute des cas où les choses fonctionnent bien et d'autres où elles ne fonctionnent pas. Nous vous encourageons à fournir des commentaires lorsque vous utilisez OpenDevin pour aider l'équipe de développement et, peut-être plus important encore, créer un corpus ouvert d'exemples de formation pour les agents de codage -- Partagez-OpenDevin !
## 📝 Comment Fournir des Commentaires
Fournir des commentaires est simple ! Lorsque vous utilisez OpenDevin, vous pouvez appuyer sur le bouton de pouce vers le haut ou vers le bas à n'importe quel moment de votre interaction. Vous serez invité à fournir votre adresse email (par exemple, afin que nous puissions vous contacter si nous voulons poser des questions de suivi), et vous pouvez choisir si vous souhaitez fournir des commentaires publiquement ou en privé.
<iframe width="560" height="315" src="https://www.youtube.com/embed/5rFx-StMVV0?si=svo7xzp6LhGK_GXr" title="Lecteur vidéo YouTube" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
## 📜 Licence de Données et Confidentialité
* Les données **publiques** seront distribuées sous la licence MIT, comme OpenDevin lui-même, et pourront être utilisées par la communauté pour former et tester des modèles. Évidemment, les commentaires que vous pouvez rendre publics seront plus précieux pour la communauté dans son ensemble, donc lorsque vous ne traitez pas d'informations sensibles, nous vous encourageons à choisir cette option !
* Les données **privées** ne seront partagées qu'avec l'équipe OpenDevin dans le but d'améliorer OpenDevin.
@@ -0,0 +1,111 @@
---
sidebar_position: 1
---
# 💻 OpenDevin
OpenDevin est un **ingénieur logiciel IA autonome** capable d'exécuter des tâches d'ingénierie complexes et de collaborer activement avec les utilisateurs sur des projets de développement logiciel.
Ce projet est entièrement open-source, vous pouvez donc l'utiliser et le modifier comme bon vous semble.
:::tip
Explorez le code source d'OpenDevin sur [GitHub](https://github.com/OpenDevin/OpenDevin) ou rejoignez l'une de nos communautés !
<a href="https://github.com/OpenDevin/OpenDevin/graphs/contributors">
<img
src="https://img.shields.io/github/contributors/opendevin/opendevin?style=for-the-badge"
alt="Contributors"
/>
</a>
<a href="https://github.com/OpenDevin/OpenDevin/network/members">
<img
src="https://img.shields.io/github/forks/opendevin/opendevin?style=for-the-badge"
alt="Forks"
/>
</a>
<a href="https://github.com/OpenDevin/OpenDevin/stargazers">
<img
src="https://img.shields.io/github/stars/opendevin/opendevin?style=for-the-badge"
alt="Stargazers"
/>
</a>
<a href="https://github.com/OpenDevin/OpenDevin/issues">
<img
src="https://img.shields.io/github/issues/opendevin/opendevin?style=for-the-badge"
alt="Issues"
/>
</a>
<br></br>
<a href="https://github.com/OpenDevin/OpenDevin/blob/main/LICENSE">
<img
src="https://img.shields.io/github/license/opendevin/opendevin?style=for-the-badge"
alt="MIT License"
/>
</a>
<br></br>
<a href="https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA">
<img
src="https://img.shields.io/badge/Slack-Join%20Us-red?logo=slack&logoColor=white&style=for-the-badge"
alt="Join our Slack community"
/>
</a>
<a href="https://discord.gg/ESHStjSjD4">
<img
src="https://img.shields.io/badge/Discord-Join%20Us-purple?logo=discord&logoColor=white&style=for-the-badge"
alt="Join our Discord community"
/>
</a>
:::
## 🛠️ Pour commencer
La manière la plus simple d'exécuter OpenDevin est à l'intérieur d'un conteneur Docker. Il fonctionne mieux avec la version la plus récente de Docker, `26.0.0`.
Vous devez utiliser Linux, Mac OS ou WSL sur Windows.
Pour démarrer OpenDevin dans un conteneur docker, exécutez les commandes suivantes dans votre terminal :
:::warning
Lorsque vous exécutez la commande suivante, les fichiers dans `./workspace` peuvent être modifiés ou supprimés.
:::
```bash
WORKSPACE_BASE=$(pwd)/workspace
docker run -it \
--pull=always \
-e SANDBOX_USER_ID=$(id -u) \
-e PERSIST_SANDBOX="true" \
-e SSH_PASSWORD="make something up here" \
-e WORKSPACE_MOUNT_PATH=$WORKSPACE_BASE \
-v $WORKSPACE_BASE:/opt/workspace_base \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 3000:3000 \
--add-host host.docker.internal:host-gateway \
--name opendevin-app-$(date +%Y%m%d%H%M%S) \
ghcr.io/opendevin/opendevin:0.6
```
Vous trouverez OpenDevin fonctionnant à l'adresse [http://localhost:3000](http://localhost:3000) avec accès à `./workspace`. Pour qu'OpenDevin fonctionne sur votre code, placez-le dans `./workspace`.
OpenDevin n'aura accès qu'à ce dossier de workspace. Le reste de votre système ne sera pas affecté car il s'exécute dans un bac à sable sécurisé de docker.
:::tip
Si vous souhaitez utiliser la version **(instable !)** la plus récente, vous pouvez utiliser `ghcr.io/opendevin/opendevin:main` comme image (dernière ligne).
:::
Pour le workflow de développement, consultez [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md).
Avez-vous des problèmes ? Consultez notre [Guide de dépannage](https://opendevin.github.io/OpenDevin/modules/usage/troubleshooting).
:::warning
OpenDevin est actuellement en cours de développement, mais vous pouvez déjà exécuter la version alpha pour voir le système de bout en bout en action.
:::
[contributors-shield]: https://img.shields.io/github/contributors/opendevin/opendevin?style=for-the-badge
[contributors-url]: https://github.com/OpenDevin/OpenDevin/graphs/contributors
[forks-shield]: https://img.shields.io/github/forks/opendevin/opendevin?style=for-the-badge
[forks-url]: https://github.com/OpenDevin/OpenDevin/network/members
[stars-shield]: https://img.shields.io/github/stars/opendevin/opendevin?style=for-the-badge
[stars-url]: https://github.com/OpenDevin/OpenDevin/stargazers
[issues-shield]: https://img.shields.io/github/issues/opendevin/opendevin?style=for-the-badge
[issues-url]: https://github.com/OpenDevin/OpenDevin/issues
[license-shield]: https://img.shields.io/github/license/opendevin/opendevin?style=for-the-badge
[license-url]: https://github.com/OpenDevin/OpenDevin/blob/main/LICENSE
@@ -0,0 +1,37 @@
# Azure OpenAI LLM
## Complétion
OpenDevin utilise LiteLLM pour les appels de complétion. Vous pouvez trouver leur documentation sur Azure [ici](https://docs.litellm.ai/docs/providers/azure)
### Configurations openai Azure
Lors de l'exécution de l'image Docker OpenDevin, vous devrez définir les variables d'environnement suivantes en utilisant `-e` :
```
LLM_BASE_URL="<azure-api-base-url>" # e.g. "https://openai-gpt-4-test-v-1.openai.azure.com/"
LLM_API_KEY="<azure-api-key>"
LLM_MODEL="azure/<your-gpt-deployment-name>"
LLM_API_VERSION = "<api-version>" # e.g. "2024-02-15-preview"
```
:::note
Vous pouvez trouver le nom de votre déploiement ChatGPT sur la page des déploiements sur Azure. Par défaut ou initialement, il pourrait être le même que le nom du modèle de chat (par exemple 'GPT4-1106-preview'), mais il n'est pas obligé de l'être. Exécutez OpenDevin, et une fois chargé dans le navigateur, allez dans Paramètres et définissez le modèle comme suit : "azure/&lt;your-actual-gpt-deployment-name&gt;". Si ce n'est pas dans la liste, entrez votre propre texte et enregistrez-le.
:::
## Embeddings
OpenDevin utilise llama-index pour les embeddings. Vous pouvez trouver leur documentation sur Azure [ici](https://docs.llamaindex.ai/en/stable/api_reference/embeddings/azure_openai/)
### Configurations openai Azure
Le modèle utilisé pour les embeddings Azure OpenAI est "text-embedding-ada-002".
Vous avez besoin du nom de déploiement correct pour ce modèle dans votre compte Azure.
Lors de l'exécution d'OpenDevin dans Docker, définissez les variables d'environnement suivantes en utilisant `-e` :
```
LLM_EMBEDDING_MODEL="azureopenai"
LLM_EMBEDDING_DEPLOYMENT_NAME = "<your-embedding-deployment-name>" # e.g. "TextEmbedding...<etc>"
LLM_API_VERSION = "<api-version>" # e.g. "2024-02-15-preview"
```
@@ -0,0 +1,28 @@
# Google Gemini/Vertex LLM
## Complétion
OpenDevin utilise LiteLLM pour les appels de complétion. Les ressources suivantes sont pertinentes pour utiliser OpenDevin avec les LLMs de Google :
- [Gemini - Google AI Studio](https://docs.litellm.ai/docs/providers/gemini)
- [VertexAI - Google Cloud Platform](https://docs.litellm.ai/docs/providers/vertex)
### Configurations de Gemini - Google AI Studio
Pour utiliser Gemini via Google AI Studio lors de l'exécution de l'image Docker d'OpenDevin, vous devez définir les variables d'environnement suivantes en utilisant `-e` :
```
GEMINI_API_KEY="<votre-cle-api-google>"
LLM_MODEL="gemini/gemini-1.5-pro"
```
### Configurations de Vertex AI - Google Cloud Platform
Pour utiliser Vertex AI via Google Cloud Platform lors de l'exécution de l'image Docker d'OpenDevin, vous devez définir les variables d'environnement suivantes en utilisant `-e` :
```
GOOGLE_APPLICATION_CREDENTIALS="<dump-json-du-compte-de-service-gcp-json>"
VERTEXAI_PROJECT="<votre-id-de-projet-gcp>"
VERTEXAI_LOCATION="<votre-localisation-gcp>"
LLM_MODEL="vertex_ai/<modele-llm-desire>"
```
@@ -0,0 +1,44 @@
---
sidebar_position: 2
---
# 🤖 Backends LLM
OpenDevin peut fonctionner avec n'importe quel backend LLM.
Pour une liste complète des fournisseurs et des modèles LM disponibles, veuillez consulter la
[documentation litellm](https://docs.litellm.ai/docs/providers).
:::warning
OpenDevin émettra de nombreuses invitations au LLM que vous configurez. La plupart de ces LLM coûtent de l'argent -- assurez-vous de définir des limites de dépenses et de surveiller l'utilisation.
:::
La variable d'environnement `LLM_MODEL` contrôle le modèle utilisé dans les interactions programmatiques.
Mais en utilisant l'interface utilisateur OpenDevin, vous devrez choisir votre modèle dans la fenêtre des paramètres (la roue dentée en bas à gauche).
Les variables d'environnement suivantes peuvent être nécessaires pour certains LLM :
- `LLM_API_KEY`
- `LLM_BASE_URL`
- `LLM_EMBEDDING_MODEL`
- `LLM_EMBEDDING_DEPLOYMENT_NAME`
- `LLM_API_VERSION`
Nous avons quelques guides pour exécuter OpenDevin avec des fournisseurs de modèles spécifiques :
- [ollama](llms/localLLMs)
- [Azure](llms/azureLLMs)
Si vous utilisez un autre fournisseur, nous vous encourageons à ouvrir une PR pour partager votre configuration !
## Remarque sur les modèles alternatifs
Les meilleurs modèles sont GPT-4 et Claude 3. Les modèles locaux et open source actuels ne sont pas aussi puissants.
Lors de l'utilisation d'un modèle alternatif, vous pouvez constater des temps d'attente prolongés entre les messages,
des réponses de mauvaise qualité ou des erreurs sur des JSON mal formés. OpenDevin
ne peut être aussi puissant que les modèles qui le pilotent -- heureusement, les membres de notre équipe travaillent activement à la construction de meilleurs modèles open source !
## Réessais d'API et limites de taux
Certains LLM ont des limites de taux et peuvent nécessiter des réessais. OpenDevin réessaiera automatiquement les demandes s'il reçoit une erreur 429 ou une erreur de connexion API.
Vous pouvez définir les variables d'environnement `LLM_NUM_RETRIES`, `LLM_RETRY_MIN_WAIT`, `LLM_RETRY_MAX_WAIT` pour contrôler le nombre de réessais et le temps entre les réessais.
Par défaut, `LLM_NUM_RETRIES` est 5 et `LLM_RETRY_MIN_WAIT`, `LLM_RETRY_MAX_WAIT` sont respectivement de 3 secondes et 60 secondes.
@@ -0,0 +1,141 @@
# LLM Local avec Ollama
Assurez-vous que le serveur Ollama est en cours d'exécution.
Pour des instructions détaillées de démarrage, consultez [ici](https://github.com/ollama/ollama)
Ce guide suppose que vous avez démarré ollama avec `ollama serve`. Si vous exécutez ollama différemment (par exemple, à l'intérieur de docker), les instructions pourraient devoir être modifiées. Veuillez noter que si vous utilisez WSL, la configuration par défaut de ollama bloque les requêtes des conteneurs docker. Voir [ici](#configuring-ollama-service-fr).
## Télécharger des modèles
Les noms des modèles Ollama peuvent être trouvés [ici](https://ollama.com/library). Pour un petit exemple, vous pouvez utiliser
le modèle `codellama:7b`. Des modèles plus grands offriront généralement de meilleures performances.
```bash
ollama pull codellama:7b
```
vous pouvez vérifier quels modèles vous avez téléchargés de cette manière :
```bash
~$ ollama list
NAME ID SIZE MODIFIED
codellama:7b 8fdf8f752f6e 3.8 GB 6 weeks ago
mistral:7b-instruct-v0.2-q4_K_M eb14864c7427 4.4 GB 2 weeks ago
starcoder2:latest f67ae0f64584 1.7 GB 19 hours ago
```
## Démarrer OpenDevin
### Docker
Utilisez les instructions [ici](../intro) pour démarrer OpenDevin en utilisant Docker.
Mais lors de l'exécution de `docker run`, vous devrez ajouter quelques arguments supplémentaires :
```bash
--add-host host.docker.internal:host-gateway \
-e LLM_API_KEY="ollama" \
-e LLM_BASE_URL="http://host.docker.internal:11434" \
```
Par exemple :
```bash
# Le répertoire que vous souhaitez qu'OpenDevin modifie. DOIT être un chemin absolu !
export WORKSPACE_BASE=$(pwd)/workspace
docker run \
-it \
--pull=always \
--add-host host.docker.internal:host-gateway \
-e SANDBOX_USER_ID=$(id -u) \
-e LLM_API_KEY="ollama" \
-e LLM_BASE_URL="http://host.docker.internal:11434" \
-e WORKSPACE_MOUNT_PATH=$WORKSPACE_BASE \
-v $WORKSPACE_BASE:/opt/workspace_base \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 3000:3000 \
ghcr.io/opendevin/opendevin:main
```
Vous devriez maintenant pouvoir vous connecter à `http://localhost:3000/`
### Compiler à partir des sources
Utilisez les instructions dans [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md) pour compiler OpenDevin.
Assurez-vous que `config.toml` soit présent en exécutant `make setup-config` qui en créera un pour vous. Dans `config.toml`, saisissez les éléments suivants :
```
LLM_MODEL="ollama/codellama:7b"
LLM_API_KEY="ollama"
LLM_EMBEDDING_MODEL="local"
LLM_BASE_URL="http://localhost:11434"
WORKSPACE_BASE="./workspace"
WORKSPACE_DIR="$(pwd)/workspace"
```
Remplacez `LLM_MODEL` par celui de votre choix si nécessaire.
Fini ! Vous pouvez maintenant démarrer Devin avec : `make run` sans Docker. Vous devriez maintenant pouvoir vous connecter à `http://localhost:3000/`
## Sélection de votre modèle
Dans l'interface OpenDevin, cliquez sur l'icône des paramètres en bas à gauche.
Ensuite, dans l'entrée `Model`, saisissez `ollama/codellama:7b`, ou le nom du modèle que vous avez téléchargé précédemment.
S'il n'apparaît pas dans un menu déroulant, ce n'est pas grave, tapez-le simplement. Cliquez sur Enregistrer lorsque vous avez terminé.
Et maintenant, vous êtes prêt à démarrer !
## Configuration du service ollama (WSL){#configuring-ollama-service-fr}
La configuration par défaut pour ollama sous WSL ne sert que localhost. Cela signifie que vous ne pouvez pas l'atteindre depuis un conteneur docker, par exemple, il ne fonctionnera pas avec OpenDevin. Testons d'abord que ollama est en cours d'exécution correctement.
```bash
ollama list # obtenir la liste des modèles installés
curl http://localhost:11434/api/generate -d '{"model":"[NAME]","prompt":"hi"}'
#ex. curl http://localhost:11434/api/generate -d '{"model":"codellama:7b","prompt":"hi"}'
#ex. curl http://localhost:11434/api/generate -d '{"model":"codellama","prompt":"hi"}' #le tag est optionnel s'il n'y en a qu'un seul
```
Une fois cela fait, testez qu'il accepte les requêtes "externes", comme celles provenant d'un conteneur docker.
```bash
docker ps # obtenir la liste des conteneurs docker en cours d'exécution, pour un test le plus précis choisissez le conteneur de sandbox open devin.
docker exec [CONTAINER ID] curl http://host.docker.internal:11434/api/generate -d '{"model":"[NAME]","prompt":"hi"}'
#ex. docker exec cd9cc82f7a11 curl http://host.docker.internal:11434/api/generate -d '{"model":"codellama","prompt":"hi"}'
```
## Correction
Maintenant faisons en sorte que cela fonctionne. Modifiez /etc/systemd/system/ollama.service avec les privilèges sudo. (Le chemin peut varier selon la distribution Linux)
```bash
sudo vi /etc/systemd/system/ollama.service
```
ou
```bash
sudo nano /etc/systemd/system/ollama.service
```
Dans la section [Service], ajoutez ces lignes
```
Environment="OLLAMA_HOST=0.0.0.0:11434"
Environment="OLLAMA_ORIGINS=*"
```
Ensuite, sauvegardez, rechargez la configuration et redémarrez le service.
```bash
sudo systemctl daemon-reload
sudo systemctl restart ollama
```
Enfin, testez que ollama est accessible depuis le conteneur
```bash
ollama list # obtenir la liste des modèles installés
docker ps # obtenir la liste des conteneurs docker en cours d'exécution, pour un test le plus précis choisissez le conteneur de sandbox open devin.
docker exec [CONTAINER ID] curl http://host.docker.internal:11434/api/generate -d '{"model":"[NAME]","prompt":"hi"}'
```
@@ -0,0 +1,207 @@
---
sidebar_position: 5
---
# 🚧 Dépannage
Il existe certains messages d'erreur qui sont souvent signalés par les utilisateurs.
Nous essaierons de rendre le processus d'installation plus facile et ces messages d'erreur
mieux à l'avenir. Mais pour l'instant, vous pouvez rechercher votre message d'erreur ci-dessous et voir s'il existe des solutions de contournement.
Pour chacun de ces messages d'erreur, **il existe un problème existant**. Veuillez ne pas
ouvrir un nouveau problème - commentez simplement dessus.
Si vous trouvez plus d'informations ou une solution de contournement pour l'un de ces problèmes, veuillez ouvrir un *PR* pour ajouter des détails à ce fichier.
:::tip
Si vous utilisez Windows et que vous rencontrez des problèmes, consultez notre [guide pour les utilisateurs de Windows (WSL)](troubleshooting/windows).
:::
## Impossible de se connecter à Docker
[Problème GitHub](https://github.com/OpenDevin/OpenDevin/issues/1226)
### Symptômes
```bash
Erreur lors de la création du contrôleur. Veuillez vérifier que Docker est en cours d'exécution et visitez `https://opendevin.github.io/OpenDevin/modules/usage/troubleshooting` pour plus d'informations sur le débogage.
```
```bash
docker.errors.DockerException: Erreur lors de la récupération de la version de l'API du serveur : ('Connection aborted.', FileNotFoundError(2, 'Aucun fichier ou répertoire de ce type'))
```
### Détails
OpenDevin utilise un conteneur Docker pour effectuer son travail en toute sécurité, sans risquer de briser votre machine.
### Solutions de contournement
* Exécutez `docker ps` pour vous assurer que docker est en cours d'exécution
* Assurez-vous que vous n'avez pas besoin de `sudo` pour exécuter docker [voir ici](https://www.baeldung.com/linux/docker-run-without-sudo)
* Si vous êtes sur un Mac, vérifiez les [exigences en matière d'autorisations](https://docs.docker.com/desktop/mac/permission-requirements/) et envisagez particulièrement d'activer l'option `Allow the default Docker socket to be used` sous `Settings > Advanced` dans Docker Desktop.
* De plus, mettez à jour Docker vers la dernière version sous `Check for Updates`
## Impossible de se connecter à la boîte SSH
[Problème GitHub](https://github.com/OpenDevin/OpenDevin/issues/1156)
### Symptômes
```python
self.shell = DockerSSHBox(
...
pexpect.pxssh.ExceptionPxssh: Impossible d'établir une connexion avec l'hôte
```
### Détails
Par défaut, OpenDevin se connecte à un conteneur en cours d'exécution via SSH. Sur certaines machines,
en particulier Windows, cela semble échouer.
### Solutions de contournement
* Redémarrez votre ordinateur (parfois cela fonctionne)
* Assurez-vous d'avoir les dernières versions de WSL et Docker
* Vérifiez que votre distribution dans WSL est également à jour
* Essayez [ce guide de réinstallation](https://github.com/OpenDevin/OpenDevin/issues/1156#issuecomment-2064549427)
## Impossible de se connecter à LLM
[Problème GitHub](https://github.com/OpenDevin/OpenDevin/issues/1208)
### Symptômes
```python
File "/app/.venv/lib/python3.12/site-packages/openai/_exceptions.py", line 81, in __init__
super().__init__(message, response.request, body=body)
^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'request'
```
### Détails
[Problèmes GitHub](https://github.com/OpenDevin/OpenDevin/issues?q=is%3Aissue+is%3Aopen+404)
Cela se produit généralement avec les configurations de LLM *locales*, lorsque OpenDevin ne parvient pas à se connecter au serveur LLM.
Consultez notre guide pour [LLMs locaux](llms/localLLMs) pour plus d'informations.
### Solutions de contournement
* Vérifiez votre `base_url` dans votre config.toml (si elle existe) sous la section "llm"
* Vérifiez que ollama (ou tout autre LLM que vous utilisez) fonctionne correctement
* Assurez-vous d'utiliser `--add-host host.docker.internal:host-gateway` lorsque vous utilisez Docker
## `404 Ressource non trouvée`
### Symptômes
```python
Traceback (most recent call last):
File "/app/.venv/lib/python3.12/site-packages/litellm/llms/openai.py", line 414, in completion
raise e
File "/app/.venv/lib/python3.12/site-packages/litellm/llms/openai.py", line 373, in completion
response = openai_client.chat.completions.create(**data, timeout=timeout) # type: ignore
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_utils/_utils.py", line 277, in wrapper
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/resources/chat/completions.py", line 579, in create
return self._post(
^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1232, in post
return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 921, in request
return self._request(
^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1012, in _request
raise self._make_status_error_from_response(err.response) from None
openai.NotFoundError: Code d'erreur : 404 - {'error': {'code': '404', 'message': 'Ressource non trouvée'}}
```
### Détails
Cela se produit lorsque LiteLLM (notre bibliothèque pour se connecter à différents fournisseurs de LLM) ne parvient pas à trouver
le point de terminaison API avec lequel vous essayez de vous connecter. Cela arrive le plus souvent aux utilisateurs de Azure ou ollama.
### Solutions de contournement
* Vérifiez que vous avez correctement défini `LLM_BASE_URL`
* Vérifiez que le modèle est correctement défini, en fonction des [docs de LiteLLM](https://docs.litellm.ai/docs/providers)
* Si vous êtes en cours d'exécution dans l'interface utilisateur, assurez-vous de définir le `model` dans le modal des paramètres
* Si vous êtes en cours d'exécution sans interface (via main.py), assurez-vous de définir `LLM_MODEL` dans votre env/config
* Assurez-vous de suivre les instructions spéciales de votre fournisseur de LLM
* [ollama](/fr/modules/usage/llms/localLLMs)
* [Azure](/fr/modules/usage/llms/azureLLMs)
* [Google](/fr/modules/usage/llms/googleLLMs)
* Assurez-vous que votre clé API est correcte
* Voyez si vous pouvez vous connecter au LLM en utilisant `curl`
* Essayez de [vous connecter via LiteLLM directement](https://github.com/BerriAI/litellm) pour tester votre configuration
## `make build` bloqué sur les installations de packages
### Symptômes
Installation de package bloquée sur `En attente...` sans aucun message d'erreur :
```bash
Opérations de package : 286 installations, 0 mises à jour, 0 suppressions
- Installation de certifi (2024.2.2) : En attente...
- Installation de h11 (0.14.0) : En attente...
- Installation de idna (3.7) : En attente...
- Installation de sniffio (1.3.1) : En attente...
- Installation de typing-extensions (4.11.0) : En attente...
```
### Détails
Dans de rares cas, `make build` peut sembler bloqué sur les installations de packages
sans aucun message d'erreur.
### Solutions de contournement
* Le gestionnaire de packages Poetry peut manquer d'un paramètre de configuration concernant
l'emplacement où doivent être recherchées les informations d'identification (keyring).
### Solution de contournement
Tout d'abord, vérifiez avec `env` si une valeur pour `PYTHON_KEYRING_BACKEND` existe.
Sinon, exécutez la commande ci-dessous pour la définir à une valeur connue et réessayez la construction :
```bash
export PYTHON_KEYRING_BACKEND=keyring.backends.null.Keyring
```
## Les sessions ne sont pas restaurées
### Symptômes
OpenDevin demande généralement s'il faut reprendre ou commencer une nouvelle session lors de l'ouverture de l'interface utilisateur.
Mais cliquer sur "Reprendre" démarre toujours une toute nouvelle discussion.
### Détails
Avec une installation standard à ce jour, les données de session sont stockées en mémoire.
Actuellement, si le service OpenDevin est redémarré, les sessions précédentes deviennent
invalides (un nouveau secret est généré) et donc non récupérables.
### Solutions de contournement
* Modifiez la configuration pour rendre les sessions persistantes en éditant le fichier `config.toml`
(dans le dossier racine d'OpenDevin) en spécifiant un `file_store` et un
`file_store_path` absolu :
```toml
file_store="local"
file_store_path="/absolute/path/to/opendevin/cache/directory"
```
* Ajoutez un secret jwt fixe dans votre .bashrc, comme ci-dessous, afin que les id de session précédents
restent acceptés.
```bash
EXPORT JWT_SECRET=A_CONST_VALUE
```
@@ -0,0 +1,76 @@
# Notes pour les utilisateurs de Windows et WSL
OpenDevin ne supporte Windows que via [WSL](https://learn.microsoft.com/en-us/windows/wsl/install).
Veuillez vous assurer de lancer toutes les commandes à l'intérieur de votre terminal WSL.
## Dépannage
### Erreur : 'docker' n'a pas pu être trouvé dans cette distribution WSL 2.
Si vous utilisez Docker Desktop, assurez-vous de le démarrer avant d'exécuter toute commande docker depuis l'intérieur de WSL.
Docker doit également avoir l'option d'intégration WSL activée.
### Recommandation : Ne pas exécuter en tant qu'utilisateur root
Pour des raisons de sécurité, il est fortement recommandé de ne pas exécuter OpenDevin en tant qu'utilisateur root, mais en tant qu'utilisateur avec un UID non nul.
De plus, les sandboxes persistants ne seront pas pris en charge lors de l'exécution en tant que root et un message approprié pourrait apparaître lors du démarrage d'OpenDevin.
Références :
* [Pourquoi il est mauvais de se connecter en tant que root](https://askubuntu.com/questions/16178/why-is-it-bad-to-log-in-as-root)
* [Définir l'utilisateur par défaut dans WSL](https://www.tenforums.com/tutorials/128152-set-default-user-windows-subsystem-linux-distro-windows-10-a.html#option2)
Astuce pour la 2e référence : pour les utilisateurs d'Ubuntu, la commande pourrait en fait être "ubuntupreview" au lieu de "ubuntu".
### Échec de la création de l'utilisateur opendevin
Si vous rencontrez l'erreur suivante lors de l'installation :
```sh
Exception: Failed to create opendevin user in sandbox: 'useradd: UID 0 is not unique'
```
Vous pouvez la résoudre en exécutant :
```sh
export SANDBOX_USER_ID=1000
```
### Installation de Poetry
* Si vous rencontrez des problèmes pour exécuter Poetry même après l'avoir installé pendant le processus de construction, il peut être nécessaire d'ajouter son chemin binaire à votre environnement :
```sh
export PATH="$HOME/.local/bin:$PATH"
```
* Si `make build` s'arrête avec une erreur telle que :
```sh
ModuleNotFoundError: no module named <module-name>
```
Cela pourrait être un problème avec le cache de Poetry.
Essayez d'exécuter ces 2 commandes l'une après l'autre :
```sh
rm -r ~/.cache/pypoetry
make build
```
### L'objet NoneType n'a pas d'attribut 'request'
Si vous rencontrez des problèmes liés au réseau, tels que `NoneType object has no attribute 'request'` lors de l'exécution de `make run`, il peut être nécessaire de configurer vos paramètres réseau WSL2. Suivez ces étapes :
* Ouvrez ou créez le fichier `.wslconfig` situé à `C:\Users\%username%\.wslconfig` sur votre machine hôte Windows.
* Ajoutez la configuration suivante au fichier `.wslconfig` :
```sh
[wsl2]
networkingMode=mirrored
localhostForwarding=true
```
* Enregistrez le fichier `.wslconfig`.
* Redémarrez WSL2 complètement en quittant toute instance WSL2 en cours d'exécution et en exécutant la commande `wsl --shutdown` dans votre invite de commande ou terminal.
* Après avoir redémarré WSL, essayez d'exécuter `make run` à nouveau.
Le problème réseau devrait être résolu.
@@ -0,0 +1,26 @@
{
"title": {
"message": "OpenDevin",
"description": "The title in the navbar"
},
"logo.alt": {
"message": "OpenDevin",
"description": "The alt text of navbar logo"
},
"item.label.Docs": {
"message": "Docs",
"description": "Navbar item with label Docs"
},
"item.label.Codebase": {
"message": "Codebase",
"description": "Navbar item with label Codebase"
},
"item.label.FAQ": {
"message": "FAQ",
"description": "Navbar item with label FAQ"
},
"item.label.GitHub": {
"message": "GitHub",
"description": "Navbar item with label GitHub"
}
}
+406
View File
@@ -0,0 +1,406 @@
{
"footer.title": {
"message": "OpenDevin"
},
"footer.docs": {
"message": "文档"
},
"footer.community": {
"message": "社区"
},
"footer.copyright": {
"message": "版权所有 © {year} OpenDevin"
},
"faq.title": {
"message": "常见问题解答",
"description": "FAQ Title"
},
"faq.description": {
"message": "常见问题解答"
},
"faq.section.title.1": {
"message": "什么是OpenDevin",
"description": "First Section Title"
},
"faq.section.highlight": {
"message": "OpenDevin",
"description": "Highlight Text"
},
"faq.section.description.1": {
"message": "是一个自主的软件工程师,能够端到端地解决软件工程和网页浏览任务。它能执行数据科学查询,如 \"查找上个月OpenDevin仓库中的拉取请求数量\",还能处理软件工程任务,例如 \"请为这个文件添加测试并验证所有测试都通过,如果没有修复该文件\"。",
"description": "Description for OpenDevin"
},
"faq.section.description.2": {
"message": "同时,OpenDevin是一个代理开发者平台和社区,用于测试和评估新代理的环境。",
"description": "Further Description for OpenDevin"
},
"faq.section.title.2": {
"message": "支持",
"description": "Support Section Title"
},
"faq.section.support.answer": {
"message": "如果您发现了可能影响他人的问题,请在 {githubLink} 上提交一个 bug。如果遇到安装困难或有其他疑问,可以访问 {discordLink} 或 {slackLink} 进行提问。",
"description": "Support Answer"
},
"faq.section.title.3": {
"message": "如何使用OpenDevin解决GitHub上的问题?",
"description": "GitHub Issue Section Title"
},
"faq.section.github.steps.intro": {
"message": "要通过OpenDevin解决GitHub上的问题,您可以发送一个提示给OpenDevin,请它按照以下步骤操作:",
"description": "GitHub Steps Introduction"
},
"faq.section.github.step1": {
"message": "阅读问题 https://github.com/OpenDevin/OpenDevin/issues/1611",
"description": "GitHub Step 1"
},
"faq.section.github.step2": {
"message": "克隆仓库并创建新分支",
"description": "GitHub Step 2"
},
"faq.section.github.step3": {
"message": "根据问题描述中的说明,修改文件以解决问题",
"description": "GitHub Step 3"
},
"faq.section.github.step4": {
"message": "使用GITHUB_TOKEN环境变量将结果推送到GitHub",
"description": "GitHub Step 4"
},
"faq.section.github.step5": {
"message": "告诉我需要前往的链接来提交拉取请求",
"description": "GitHub Step 5"
},
"faq.section.github.steps.preRun": {
"message": "在运行OpenDevin之前,您可以:",
"description": "GitHub Steps Pre-Run"
},
"faq.section.github.steps.tokenInfo": {
"message": "其中XXX是您创建的一个具有对OpenDevin仓库写权限的GitHub令牌。如果您的写入权限不足,请将其更改为:",
"description": "GitHub Steps Token Info"
},
"faq.section.github.steps.usernameInfo": {
"message": "其中USERNAME是您的GitHub用户名。",
"description": "GitHub Steps Username Info"
},
"faq.section.title.4": {
"message": "OpenDevin与Devin有何不同?",
"description": "Devin Section Title"
},
"faq.section.devin.linkText": {
"message": "Devin",
"description": "Devin Link Text"
},
"faq.section.devin.description": {
"message": "是由Cognition Inc.开发的商业产品,它最初为OpenDevin提供了灵感。它们都旨在擅长解决软件工程任务,但您可以下载、使用和修改OpenDevin,而Devin只能通过Cognition网站进行访问。此外,OpenDevin已超越最初的灵感,并成为一个面向代理开发者的社区驱动生态系统,在这里我们欢迎您加入并",
"description": "Devin Description"
},
"faq.section.devin.contribute": {
"message": "贡献",
"description": "Contribute Link"
},
"faq.section.title.5": {
"message": "OpenDevin与ChatGPT有何不同?",
"description": "ChatGPT Section Title"
},
"faq.section.chatgpt.description": {
"message": "您可以通过网络访问ChatGPT,它不与本地文件交互,并且其执行代码的能力有限。因此,它可以编写代码,但测试或执行起来可能不太容易。",
"description": "ChatGPT Description"
},
"homepage.description": {
"message": "使用AI生成代码的软件工程工具。",
"description": "The homepage description"
},
"homepage.getStarted": {
"message": "开始使用"
},
"welcome.message": {
"message": "欢迎来到OpenDevin,这是一个开源自主AI软件工程师,能够执行复杂的工程任务,并积极参与用户在软件开发项目中的协作。"
},
"theme.ErrorPageContent.title": {
"message": "页面已崩溃。",
"description": "The title of the fallback page when the page crashed"
},
"theme.BackToTopButton.buttonAriaLabel": {
"message": "返回顶部",
"description": "The ARIA label for the back to top button"
},
"theme.blog.archive.title": {
"message": "历史博文",
"description": "The page & hero title of the blog archive page"
},
"theme.blog.archive.description": {
"message": "历史博文",
"description": "The page & hero description of the blog archive page"
},
"theme.blog.paginator.navAriaLabel": {
"message": "博文列表分页导航",
"description": "The ARIA label for the blog pagination"
},
"theme.blog.paginator.newerEntries": {
"message": "较新的博文",
"description": "The label used to navigate to the newer blog posts page (previous page)"
},
"theme.blog.paginator.olderEntries": {
"message": "较旧的博文",
"description": "The label used to navigate to the older blog posts page (next page)"
},
"theme.blog.post.paginator.navAriaLabel": {
"message": "博文分页导航",
"description": "The ARIA label for the blog posts pagination"
},
"theme.blog.post.paginator.newerPost": {
"message": "较新一篇",
"description": "The blog post button label to navigate to the newer/previous post"
},
"theme.blog.post.paginator.olderPost": {
"message": "较旧一篇",
"description": "The blog post button label to navigate to the older/next post"
},
"theme.blog.post.plurals": {
"message": "{count} 篇博文",
"description": "Pluralized label for \"{count} posts\". Use as much plural forms (separated by \"|\") as your language support (see https://www.unicode.org/cldr/cldr-aux/charts/34/supplemental/language_plural_rules.html)"
},
"theme.blog.tagTitle": {
"message": "{nPosts} 含有标签「{tagName}」",
"description": "The title of the page for a blog tag"
},
"theme.tags.tagsPageLink": {
"message": "查看所有标签",
"description": "The label of the link targeting the tag list page"
},
"theme.colorToggle.ariaLabel": {
"message": "切换浅色/暗黑模式(当前为{mode})",
"description": "The ARIA label for the navbar color mode toggle"
},
"theme.colorToggle.ariaLabel.mode.dark": {
"message": "暗黑模式",
"description": "The name for the dark color mode"
},
"theme.colorToggle.ariaLabel.mode.light": {
"message": "浅色模式",
"description": "The name for the light color mode"
},
"theme.docs.breadcrumbs.navAriaLabel": {
"message": "页面路径",
"description": "The ARIA label for the breadcrumbs"
},
"theme.docs.DocCard.categoryDescription.plurals": {
"message": "{count} 个项目",
"description": "The default description for a category card in the generated index about how many items this category includes"
},
"theme.docs.paginator.navAriaLabel": {
"message": "文件选项卡",
"description": "The ARIA label for the docs pagination"
},
"theme.docs.paginator.previous": {
"message": "上一页",
"description": "The label used to navigate to the previous doc"
},
"theme.docs.paginator.next": {
"message": "下一页",
"description": "The label used to navigate to the next doc"
},
"theme.docs.tagDocListPageTitle.nDocsTagged": {
"message": "{count} 篇文档带有标签",
"description": "Pluralized label for \"{count} docs tagged\". Use as much plural forms (separated by \"|\") as your language support (see https://www.unicode.org/cldr/cldr-aux/charts/34/supplemental/language_plural_rules.html)"
},
"theme.docs.tagDocListPageTitle": {
"message": "{nDocsTagged}「{tagName}」",
"description": "The title of the page for a docs tag"
},
"theme.docs.versionBadge.label": {
"message": "版本:{versionLabel}"
},
"theme.docs.versions.unreleasedVersionLabel": {
"message": "此为 {siteTitle} {versionLabel} 版尚未发行的文档。",
"description": "The label used to tell the user that he's browsing an unreleased doc version"
},
"theme.docs.versions.unmaintainedVersionLabel": {
"message": "此为 {siteTitle} {versionLabel} 版的文档,现已不再积极维护。",
"description": "The label used to tell the user that he's browsing an unmaintained doc version"
},
"theme.docs.versions.latestVersionSuggestionLabel": {
"message": "最新的文档请参阅 {latestVersionLink} ({versionLabel})。",
"description": "The label used to tell the user to check the latest version"
},
"theme.docs.versions.latestVersionLinkLabel": {
"message": "最新版本",
"description": "The label used for the latest version suggestion link label"
},
"theme.common.editThisPage": {
"message": "编辑此页",
"description": "The link label to edit the current page"
},
"theme.common.headingLinkTitle": {
"message": "{heading}的直接链接",
"description": "Title for link to heading"
},
"theme.lastUpdated.atDate": {
"message": "于 {date} ",
"description": "The words used to describe on which date a page has been last updated"
},
"theme.lastUpdated.byUser": {
"message": "由 {user} ",
"description": "The words used to describe by who the page has been last updated"
},
"theme.lastUpdated.lastUpdatedAtBy": {
"message": "最后{byUser}{atDate}更新",
"description": "The sentence used to display when a page has been last updated, and by who"
},
"theme.navbar.mobileVersionsDropdown.label": {
"message": "选择版本",
"description": "The label for the navbar versions dropdown on mobile view"
},
"theme.NotFound.title": {
"message": "找不到页面",
"description": "The title of the 404 page"
},
"theme.tags.tagsListLabel": {
"message": "标签:",
"description": "The label alongside a tag list"
},
"theme.admonition.caution": {
"message": "警告",
"description": "The default label used for the Caution admonition (:::caution)"
},
"theme.admonition.danger": {
"message": "危险",
"description": "The default label used for the Danger admonition (:::danger)"
},
"theme.admonition.info": {
"message": "信息",
"description": "The default label used for the Info admonition (:::info)"
},
"theme.admonition.note": {
"message": "备注",
"description": "The default label used for the Note admonition (:::note)"
},
"theme.admonition.tip": {
"message": "提示",
"description": "The default label used for the Tip admonition (:::tip)"
},
"theme.admonition.warning": {
"message": "注意",
"description": "The default label used for the Warning admonition (:::warning)"
},
"theme.AnnouncementBar.closeButtonAriaLabel": {
"message": "关闭",
"description": "The ARIA label for close button of announcement bar"
},
"theme.blog.sidebar.navAriaLabel": {
"message": "最近博文导航",
"description": "The ARIA label for recent posts in the blog sidebar"
},
"theme.CodeBlock.copied": {
"message": "复制成功",
"description": "The copied button label on code blocks"
},
"theme.CodeBlock.copyButtonAriaLabel": {
"message": "将代码复制到剪贴板",
"description": "The ARIA label for copy code blocks button"
},
"theme.CodeBlock.copy": {
"message": "复制",
"description": "The copy button label on code blocks"
},
"theme.CodeBlock.wordWrapToggle": {
"message": "切换自动换行",
"description": "The title attribute for toggle word wrapping button of code block lines"
},
"theme.DocSidebarItem.expandCategoryAriaLabel": {
"message": "展开侧边栏分类 '{label}'",
"description": "The ARIA label to expand the sidebar category"
},
"theme.DocSidebarItem.collapseCategoryAriaLabel": {
"message": "折叠侧边栏分类 '{label}'",
"description": "The ARIA label to collapse the sidebar category"
},
"theme.NavBar.navAriaLabel": {
"message": "主导航",
"description": "The ARIA label for the main navigation"
},
"theme.navbar.mobileLanguageDropdown.label": {
"message": "选择语言",
"description": "The label for the mobile language switcher dropdown"
},
"theme.NotFound.p1": {
"message": "我们找不到您要找的页面。",
"description": "The first paragraph of the 404 page"
},
"theme.NotFound.p2": {
"message": "请联系原始链接来源网站的所有者,并告知他们链接已损坏。",
"description": "The 2nd paragraph of the 404 page"
},
"theme.TOCCollapsible.toggleButtonLabel": {
"message": "本页总览",
"description": "The label used by the button on the collapsible TOC component"
},
"theme.blog.post.readMore": {
"message": "阅读更多",
"description": "The label used in blog post item excerpts to link to full blog posts"
},
"theme.blog.post.readMoreLabel": {
"message": "阅读 {title} 的全文",
"description": "The ARIA label for the link to full blog posts from excerpts"
},
"theme.blog.post.readingTime.plurals": {
"message": "阅读需 {readingTime} 分钟",
"description": "Pluralized label for \"{readingTime} min read\". Use as much plural forms (separated by \"|\") as your language support (see https://www.unicode.org/cldr/cldr-aux/charts/34/supplemental/language_plural_rules.html)"
},
"theme.docs.breadcrumbs.home": {
"message": "主页面",
"description": "The ARIA label for the home page in the breadcrumbs"
},
"theme.docs.sidebar.collapseButtonTitle": {
"message": "收起侧边栏",
"description": "The title attribute for collapse button of doc sidebar"
},
"theme.docs.sidebar.collapseButtonAriaLabel": {
"message": "收起侧边栏",
"description": "The title attribute for collapse button of doc sidebar"
},
"theme.docs.sidebar.navAriaLabel": {
"message": "文档侧边栏",
"description": "The ARIA label for the sidebar navigation"
},
"theme.docs.sidebar.closeSidebarButtonAriaLabel": {
"message": "关闭导航栏",
"description": "The ARIA label for close button of mobile sidebar"
},
"theme.navbar.mobileSidebarSecondaryMenu.backButtonLabel": {
"message": "← 回到主菜单",
"description": "The label of the back button to return to main menu, inside the mobile navbar sidebar secondary menu (notably used to display the docs sidebar)"
},
"theme.docs.sidebar.toggleSidebarButtonAriaLabel": {
"message": "切换导航栏",
"description": "The ARIA label for hamburger menu button of mobile navigation"
},
"theme.docs.sidebar.expandButtonTitle": {
"message": "展开侧边栏",
"description": "The ARIA label and title attribute for expand button of doc sidebar"
},
"theme.docs.sidebar.expandButtonAriaLabel": {
"message": "展开侧边栏",
"description": "The ARIA label and title attribute for expand button of doc sidebar"
},
"theme.ErrorPageContent.tryAgain": {
"message": "重试",
"description": "The label of the button to try again rendering when the React error boundary captures an error"
},
"theme.common.skipToMainContent": {
"message": "跳到主要内容",
"description": "The skip to content label used for accessibility, allowing to rapidly navigate to main content with keyboard tab/enter navigation"
},
"theme.tags.tagsPageTitle": {
"message": "标签",
"description": "The title of the tag list page"
},
"theme.unlistedContent.title": {
"message": "未列出页",
"description": "The unlisted content banner title"
},
"theme.unlistedContent.message": {
"message": "此页面未列出。搜索引擎不会对其索引,只有拥有直接链接的用户才能访问。",
"description": "The unlisted content banner message"
}
}
@@ -0,0 +1,14 @@
{
"title": {
"message": "博客",
"description": "The title for the blog used in SEO"
},
"description": {
"message": "博客",
"description": "The description for the blog used in SEO"
},
"sidebar.title": {
"message": "最近文章",
"description": "The label for the left sidebar"
}
}
@@ -0,0 +1,18 @@
{
"version.label": {
"message": "Next",
"description": "The label for version current"
},
"sidebar.docsSidebar.category.🤖 LLM 支持": {
"message": "🤖 LLM 支持",
"description": "The label for category 🤖 LLM 支持 in sidebar docsSidebar"
},
"sidebar.docsSidebar.category.🚧 故障排除": {
"message": "🚧 故障排除",
"description": "The label for category 🚧 故障排除 in sidebar docsSidebar"
},
"sidebar.apiSidebar.category.Backend": {
"message": "Backend",
"description": "The label for category Backend in sidebar apiSidebar"
}
}
@@ -0,0 +1,3 @@
# Python 文档
部署后文档将会显示在这里。
@@ -0,0 +1,5 @@
{
"items": ["python/python"],
"label": "后端",
"type": "category"
}
@@ -0,0 +1,53 @@
---
sidebar_position: 7
---
# 📚 杂项
## ⭐️ 研究策略
通过 LLM 完全复制生产级应用程序是一个复杂的任务。我们的策略包含以下几个方面:
1. **核心技术研究:** 专注于基础研究,以理解和改进代码生成和处理的技术方面。
2. **专家能力:** 通过数据策划、训练方法等方式增强核心组件的有效性。
3. **任务规划:** 开发错误检测、代码库管理和优化的能力。
4. **评价:** 建立全面的评价指标,以更好地理解和改进我们的模型。
## 🚧 默认代理
- 我们当前的默认代理是 CodeActAgent,具备生成代码和处理文件的能力。我们正在开发其他代理实现,包括 [SWE Agent](https://swe-agent.com/)。您可以[在这里阅读我们当前的代理集合](./agents)。
## 🤝 如何贡献
OpenDevin 是一个社区驱动的项目,我们欢迎每个人的贡献。无论您是开发人员、研究人员,还是对用 AI 提升软件工程领域有兴趣,只要您愿意参与,我们都有很多方式可供选择:
- **代码贡献:** 帮助我们开发核心功能、前端界面或沙箱解决方案。
- **研究和评价:** 贡献您对 LLM 在软件工程领域理解的见解,参与评估模型,或提出改进建议。
- **反馈和测试:** 使用 OpenDevin 工具集,报告错误,建议功能,或提供可用性方面的反馈。
详情请查阅[此文件](https://github.com/OpenDevin/OpenDevin/blob/main/CONTRIBUTING.md)。
## 🤖 加入我们的社区
我们现在有一个 Slack 工作区,用于合作建设 OpenDevin,还设有一个 Discord 服务器,用于讨论与该项目、LLM、代理等相关的任何事情。
- [Slack 工作区](https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA)
- [Discord 服务器](https://discord.gg/ESHStjSjD4)
如果您愿意贡献,请随时加入我们的社区。让我们一起简化软件工程!
🐚 **少写代码,用 OpenDevin 做更多的事情。**
[![Star History Chart](https://api.star-history.com/svg?repos=OpenDevin/OpenDevin&type=Date)](https://star-history.com/#OpenDevin/OpenDevin&Date)
## 🛠️ 技术选型
OpenDevin 使用了一系列强大的框架和库,提供了坚实的开发基础。以下是项目中使用的关键技术:
![FastAPI](https://img.shields.io/badge/FastAPI-black?style=for-the-badge) ![uvicorn](https://img.shields.io/badge/uvicorn-black?style=for-the-badge) ![LiteLLM](https://img.shields.io/badge/LiteLLM-black?style=for-the-badge) ![Docker](https://img.shields.io/badge/Docker-black?style=for-the-badge) ![Ruff](https://img.shields.io/badge/Ruff-black?style=for-the-badge) ![MyPy](https://img.shields.io/badge/MyPy-black?style=for-the-badge) ![LlamaIndex](https://img.shields.io/badge/LlamaIndex-black?style=for-the-badge) ![React](https://img.shields.io/badge/React-black?style=for-the-badge)
请注意,这些技术选型仍在进行中,随着项目的发展,可能会添加新的技术或移除现有的技术。我们努力采用最适合、最高效的工具,以增强 OpenDevin 的能力。
## 📜 许可证
根据 MIT 许可证分发。详见[我们的许可证](https://github.com/OpenDevin/OpenDevin/blob/main/LICENSE)了解更多信息。
@@ -0,0 +1,134 @@
---
sidebar_position: 3
---
# 🧠 Agents and Capabilities
## CodeAct Agent
### 描述
该Agent实现了CodeAct的思想([论文](https://arxiv.org/abs/2402.01030)[推特](https://twitter.com/xingyaow_/status/1754556835703751087)),将LLM agents的**行为**合并到一个统一的**代码**动作空间中,以实现_简化_和_性能_(详情见论文)。
概念理念如下图所示。在每个回合,Agent可以:
1. **对话**:用自然语言与人类交流,进行澄清、确认等。
2. **CodeAct**:选择通过执行代码来完成任务
- 执行任何有效的Linux `bash`命令
- 使用[交互式Python解释器](https://ipython.org/)执行任何有效的 `Python`代码。这是通过`bash`命令模拟的,详细信息请参见插件系统。
![image](https://github.com/OpenDevin/OpenDevin/assets/38853559/92b622e3-72ad-4a61-8f41-8c040b6d5fb3)
### 插件系统
为了使CodeAct agent在仅能访问`bash`动作空间时更强大,CodeAct agent利用了OpenDevin的插件系统:
- [Jupyter插件](https://github.com/OpenDevin/OpenDevin/tree/main/opendevin/runtime/plugins/jupyter):通过bash命令实现IPython执行
- [SWE-agent工具插件](https://github.com/OpenDevin/OpenDevin/tree/main/opendevin/runtime/plugins/swe_agent_commands):为软件开发任务引入的强大bash命令行工具,由[swe-agent](https://github.com/princeton-nlp/swe-agent)提供。
### 演示
https://github.com/OpenDevin/OpenDevin/assets/38853559/f592a192-e86c-4f48-ad31-d69282d5f6ac
_CodeActAgent使用`gpt-4-turbo-2024-04-09`执行数据科学任务(线性回归)的示例_
### 动作
`Action`,
`CmdRunAction`,
`IPythonRunCellAction`,
`AgentEchoAction`,
`AgentFinishAction`,
`AgentTalkAction`
### 观测
`CmdOutputObservation`,
`IPythonRunCellObservation`,
`AgentMessageObservation`,
`UserMessageObservation`
### 方法
| 方法 | 描述 |
| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------ |
| `__init__` | 使用`llm`和一系列信息`list[Mapping[str, str]]`初始化Agent |
| `step` | 使用CodeAct Agent执行一步操作,包括收集前一步的信息并提示模型执行命令。 |
### 进行中的工作 & 下一步
[] 支持Web浏览
[] 完成CodeAct agent提交Github PR的工作流程
## Monologue Agent
### 描述
Monologue Agent利用长短期记忆来完成任务。
长期记忆存储为LongTermMemory对象,模型使用它来搜索过去的示例。
短期记忆存储为Monologue对象,模型可以根据需要进行压缩。
### 动作
`Action`,
`NullAction`,
`CmdRunAction`,
`FileWriteAction`,
`FileReadAction`,
`BrowseURLAction`,
`GithubPushAction`,
`AgentThinkAction`
### 观测
`Observation`,
`NullObservation`,
`CmdOutputObservation`,
`FileReadObservation`,
`BrowserOutputObservation`
### 方法
| 方法 | 描述 |
| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------ |
| `__init__` | 使用长期记忆和内部独白初始化Agent |
| `_add_event` | 将事件附加到Agent的独白中,如独白过长自动与摘要一起压缩 |
| `_initialize` | 使用`INITIAL_THOUGHTS`列表为agent提供其能力的上下文以及如何导航`/workspace` |
| `step` | 通过添加最近的动作和观测修改当前状态,然后提示模型考虑其接下来的动作。 |
## Planner Agent
### 描述
Planner agent利用特殊的提示策略为解决问题创建长期计划。
在每一步中,Agent会获得其先前的动作-观测对、当前任务以及基于上一次操作提供的提示。
### 动作
`NullAction`,
`CmdRunAction`,
`BrowseURLAction`,
`GithubPushAction`,
`FileReadAction`,
`FileWriteAction`,
`AgentThinkAction`,
`AgentFinishAction`,
`AgentSummarizeAction`,
`AddTaskAction`,
`ModifyTaskAction`
### 观测
`Observation`,
`NullObservation`,
`CmdOutputObservation`,
`FileReadObservation`,
`BrowserOutputObservation`
### 方法
| 方法 | 描述 |
| -------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `__init__` | 使用`llm`初始化Agent |
| `step` | 检查当前步骤是否完成,如果是则返回`AgentFinishAction`。否则,创建计划提示并发送给模型进行推理,将结果作为下一步动作。 |
@@ -0,0 +1,49 @@
---
sidebar_position: 4
---
# 🏛️ 系统架构概览
这是系统架构的高层概览。系统分为两个主要组件:前端和后端。前端负责处理用户交互并显示结果。后端负责处理业务逻辑并执行代理。
![system_architecture.svg](/img/system_architecture.svg)
此概览简化显示了主要组件及其交互。有关后端架构的更详细视图,请参见[后端架构](#backend-architecture-zh-Hans)部分。
# 后端架构 {#backend-architecture-zh-Hans}
_**免责声明**:后端架构正在进行中,可能会有所变化。下图显示了基于图表页脚中的提交内容的当前后端架构。_
![backend_architecture.svg](/img/backend_architecture.svg)
<details>
<summary>更新此图表</summary>
<div>
后端架构图的生成部分是自动化的。
图表是使用 py2puml 工具从代码中的类型提示生成的。
然后人工审核、调整并导出为 PNG 和 SVG。
## 前提条件
- 能运行 python 环境,其中 opendevin 可以执行(根据存储库根目录中的 README.md 文件中的说明)
- 安装了 [py2puml](https://github.com/lucsorel/py2puml)
## 步骤
1. 通过从存储库根目录运行以下命令自动生成图表:
`py2puml opendevin opendevin > docs/architecture/backend_architecture.puml`
2. 在 PlantUML 编辑器中打开生成的文件,例如使用 PlantUML 扩展的 Visual Studio Code 或 [PlantText](https://www.planttext.com/)
3. 审查生成的 PUML 并对图表进行所有必要的调整(添加缺失部分、修正错误、改进定位)。
_py2puml 根据代码中的类型提示创建图表,因此缺失或不正确的类型提示可能导致图表不完整或不正确。_
4. 审查新旧图表之间的差异,并手动检查更改是否正确。
_确保不移除过去手动添加到图表中的和仍然相关的部分。_
5. 将用于生成图表的提交哈希添加到图表页脚。
6. 将图表导出为 PNG 和 SVG 文件,并替换 `docs/architecture` 目录中的现有图表。这可以通过(例如 [PlantText](https://www.planttext.com/))完成。
</div>
</details>
@@ -0,0 +1,107 @@
# 💿 如何创建自定义 Docker 沙箱
默认的 OpenDevin 沙箱包含一个[最小化 ubuntu 配置](https://github.com/OpenDevin/OpenDevin/blob/main/containers/sandbox/Dockerfile)。您的应用场景可能需要在默认状态下安装额外的软件。本指南将教您如何通过使用自定义 Docker 映像来实现这一目标。
## 环境设置
确保您能够首先通过 [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md) 运行 OpenDevin。
## 创建您的 Docker 映像
接下来,您必须创建一个自定义的 Docker 映像,该映像是基于 Debian 或 Ubuntu 的。例如,如果我们希望 OpenDevin 能够访问 "node" 可执行文件,我们可以使用以下 `Dockerfile`:
```bash
# 从最新版 ubuntu 开始
FROM ubuntu:latest
# 运行必要的更新
RUN apt-get update && apt-get install
# 安装 node
RUN apt-get install -y nodejs
```
然后构建您选择的映像,例如“custom_image”。为此可以在目录中创建文件夹并将 `Dockerfile` 放入其中,并在该目录内运行以下命令:
```bash
docker build -t custom_image .
```
这将生成一个名为 ```custom_image``` 的新映像,并使其可用于 Docker 服务引擎。
> 注意:在本文档描述的配置中,OpenDevin 将在沙箱内部以“opendevin”用户身份运行。因此,通过 Dockerfile 安装的所有包应可供系统上的所有用户使用,而不仅仅是 root 用户。
> 使用 `apt-get` 上面安装的 node 是为所有用户安装的。
## 在 config.toml 文件中指定自定义映像
在 OpenDevin 的配置通过顶层的 `config.toml` 文件发生。在 OpenDevin 目录下创建一个 ```config.toml``` 文件,并输入以下内容:
```
[core]
workspace_base="./workspace"
persist_sandbox=false
run_as_devin=true
sandbox_container_image="custom_image"
```
> 确保 `sandbox_container_image` 设置为您前一步中自定义映像的名称。
## 运行
通过运行 ```make run``` 在顶层目录下运行 OpenDevin。
导航至 ```localhost:3001``` 并检查所需依赖是否可用。
在上述示例的情况下,终端中运行 `node -v` 会输出 `v18.19.1`。
恭喜您!
## 技术解释
相关代码定义在 [ssh_box.py](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/ssh_box.py) 和 [image_agnostic_util.py](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py) 中。
特别是 ssh_box.py 检查配置对象中的 ```config.sandbox_container_image```,然后尝试使用 [get_od_sandbox_image](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L72),在 image_agnostic_util.py 定义中进行检索。
初次使用自定义映像时,该映像将不会被找到,因此将被构建(在后续运行中已构建的映像将被查找并返回)。
自定义映像是通过 `_build_sandbox_image()` 构建的,在 [image_agnostic_util.py](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L29) 中,使用您的 custom_image 作为基础,并为 OpenDevin 配置环境。例如:
```python
dockerfile_content = (
f'FROM {base_image}\n'
'RUN apt update && apt install -y openssh-server wget sudo\n'
'RUN mkdir -p -m0755 /var/run/sshd\n'
'RUN mkdir -p /opendevin && mkdir -p /opendevin/logs && chmod 777 /opendevin/logs\n'
'RUN wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"\n'
'RUN bash Miniforge3-$(uname)-$(uname -m).sh -b -p /opendevin/miniforge3\n'
'RUN bash -c ". /opendevin/miniforge3/etc/profile.d/conda.sh && conda config --set changeps1 False && conda config --append channels conda-forge"\n'
'RUN echo "export PATH=/opendevin/miniforge3/bin:$PATH" >> ~/.bashrc\n'
'RUN echo "export PATH=/opendevin/miniforge3/bin:$PATH" >> /opendevin/bash.bashrc\n'
).strip()
```
> 注意:映像名称通过 [_get_new_image_name()](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L63) 修改,并且是后续运行中搜索的修改后的名称。
## 故障排除 / 错误
### 错误:```useradd: UID 1000 is not unique```
如果在控制台输出中看到此错误,说明 OpenDevin 尝试在沙箱中以 UID 1000 创建 opendevin 用户,但该 UID 已经被映像中的其他部分使用(不知何故)。要解决这个问题,请更改 config.toml 文件中的 sandbox_user_id 字段为不同的值:
```
[core]
workspace_base="./workspace"
persist_sandbox=false
run_as_devin=true
sandbox_container_image="custom_image"
sandbox_user_id="1001"
```
### 端口使用错误
如果您看到关于端口被占用或不可用的错误,请尝试删除所有正在运行的 Docker 容器(通过运行 `docker ps` 和 `docker rm` 相关容器),然后重新运行 ```make run```。
## 讨论
对于其他问题或疑问,请加入 [Slack](https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA) 或 [Discord](https://discord.gg/ESHStjSjD4),并提问!
@@ -0,0 +1,18 @@
---
sidebar_position: 6
---
# ✅ 提供反馈
在使用 OpenDevin 时,你无疑会遇到一些情况,某些地方工作得很好,而另一些地方则可能不尽如人意。我们鼓励你在使用 OpenDevin 时提供反馈,这不仅有助于开发团队改善应用,更为重要的是,可以创建一个开放的编码代理训练样例语料库——Share-OpenDevin
## 📝 如何提供反馈
提供反馈很简单!在使用 OpenDevin 时,你可以在任意时刻按下点赞或点踩按钮。你将被要求提供你的电子邮件地址(例如,以便我们在需要进一步询问时联系你),你可以选择公开或私密地提供反馈。
<iframe width="560" height="315" src="https://www.youtube.com/embed/5rFx-StMVV0?si=svo7xzp6LhGK_GXr" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
## 📜 数据许可与隐私
* **公开** 数据将与 OpenDevin 本身一样以 MIT 许可协议发布,并可被社区用来训练和测试模型。显然,你能够公开的反馈对整个社区来说更有价值,因此当你不涉及敏感信息时,我们鼓励你选择这个选项!
* **私密** 数据将仅与 OpenDevin 团队分享,用于改进 OpenDevin。
@@ -0,0 +1,111 @@
---
sidebar_position: 1
---
# 💻 OpenDevin
OpenDevin 是一个**自主 AI 软件工程师**,能够执行复杂的工程任务,并在软件开发项目中积极与用户合作。
这个项目是完全开源的,所以你可以随意使用和修改它。
:::tip
在 [GitHub](https://github.com/OpenDevin/OpenDevin) 上探索 OpenDevin 的代码库或加入我们的社区!
<a href="https://github.com/OpenDevin/OpenDevin/graphs/contributors">
<img
src="https://img.shields.io/github/contributors/opendevin/opendevin?style=for-the-badge"
alt="Contributors"
/>
</a>
<a href="https://github.com/OpenDevin/OpenDevin/network/members">
<img
src="https://img.shields.io/github/forks/opendevin/opendevin?style=for-the-badge"
alt="Forks"
/>
</a>
<a href="https://github.com/OpenDevin/OpenDevin/stargazers">
<img
src="https://img.shields.io/github/stars/opendevin/opendevin?style=for-the-badge"
alt="Stargazers"
/>
</a>
<a href="https://github.com/OpenDevin/OpenDevin/issues">
<img
src="https://img.shields.io/github/issues/opendevin/opendevin?style=for-the-badge"
alt="Issues"
/>
</a>
<br></br>
<a href="https://github.com/OpenDevin/OpenDevin/blob/main/LICENSE">
<img
src="https://img.shields.io/github/license/opendevin/opendevin?style=for-the-badge"
alt="MIT License"
/>
</a>
<br></br>
<a href="https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA">
<img
src="https://img.shields.io/badge/Slack-Join%20Us-red?logo=slack&logoColor=white&style=for-the-badge"
alt="Join our Slack community"
/>
</a>
<a href="https://discord.gg/ESHStjSjD4">
<img
src="https://img.shields.io/badge/Discord-Join%20Us-purple?logo=discord&logoColor=white&style=for-the-badge"
alt="Join our Discord community"
/>
</a>
:::
## 🛠️ 入门指南
运行 OpenDevin 最简单的方法是在 Docker 容器中。它在 Docker 的最新版本 `26.0.0` 上运行效果最佳。
你必须使用 Linux、Mac OS 或 Windows 上的 WSL。
要在 Docker 容器中启动 OpenDevin,请在终端中运行以下命令:
:::warning
运行以下命令时,`./workspace` 中的文件可能会被修改或删除。
:::
```bash
WORKSPACE_BASE=$(pwd)/workspace
docker run -it \
--pull=always \
-e SANDBOX_USER_ID=$(id -u) \
-e PERSIST_SANDBOX="true" \
-e SSH_PASSWORD="make something up here" \
-e WORKSPACE_MOUNT_PATH=$WORKSPACE_BASE \
-v $WORKSPACE_BASE:/opt/workspace_base \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 3000:3000 \
--add-host host.docker.internal:host-gateway \
--name opendevin-app-$(date +%Y%m%d%H%M%S) \
ghcr.io/opendevin/opendevin:0.6
```
你会发现 OpenDevin 在 [http://localhost:3000](http://localhost:3000) 运行,并可以访问 `./workspace`。要让 OpenDevin 操作你的代码,请将代码放在 `./workspace` 中。
OpenDevin 只会访问这个工作区文件夹。它在一个安全的 docker 沙盒中运行,不会影响你系统的其他部分。
:::tip
如果你想使用**(不稳定!)**最新版本,可以使用 `ghcr.io/opendevin/opendevin:main` 作为镜像(最后一行)。
:::
有关开发工作流程,请参阅 [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md)。
遇到问题了吗?查看我们的 [故障排除指南](https://opendevin.github.io/OpenDevin/modules/usage/troubleshooting)。
:::warning
OpenDevin 目前正在开发中,但你已经可以运行 alpha 版本来查看端到端系统的运作情况。
:::
[contributors-shield]: https://img.shields.io/github/contributors/opendevin/opendevin?style=for-the-badge
[contributors-url]: https://github.com/OpenDevin/OpenDevin/graphs/contributors
[forks-shield]: https://img.shields.io/github/forks/opendevin/opendevin?style=for-the-badge
[forks-url]: https://github.com/OpenDevin/OpenDevin/network/members
[stars-shield]: https://img.shields.io/github/stars/opendevin/opendevin?style=for-the-badge
[stars-url]: https://github.com/OpenDevin/OpenDevin/stargazers
[issues-shield]: https://img.shields.io/github/issues/opendevin/opendevin?style=for-the-badge
[issues-url]: https://github.com/OpenDevin/OpenDevin/issues
[license-shield]: https://img.shields.io/github/license/opendevin/opendevin?style=for-the-badge
[license-url]: https://github.com/OpenDevin/OpenDevin/blob/main/LICENSE
@@ -0,0 +1,37 @@
# Azure OpenAI 大型语言模型
## 完成
OpenDevin 使用 LiteLLM 进行完成调用。你可以在 Azure 的文档中找到他们的文档 [这里](https://docs.litellm.ai/docs/providers/azure)
### Azure openai 配置
在运行 OpenDevin Docker 镜像时,你需要使用 `-e` 设置以下环境变量:
```
LLM_BASE_URL="<azure-api-base-url>" # 示例: "https://openai-gpt-4-test-v-1.openai.azure.com/"
LLM_API_KEY="<azure-api-key>"
LLM_MODEL="azure/<your-gpt-deployment-name>"
LLM_API_VERSION = "<api-version>" # 示例: "2024-02-15-preview"
```
:::note
你可以在 Azure 的部署页面找到你的 ChatGPT 部署名称。它可能与默认或最初设置的聊天模型名称相同(例如 'GPT4-1106-preview'),但不一定相同。运行 OpenDevin,当你在浏览器中加载它时,进入设置并按照上述方式设置模型: "azure/&lt;your-actual-gpt-deployment-name&gt;"。如果列表中没有,请输入你自己的文本并保存。
:::
## 嵌入
OpenDevin 使用 llama-index 进行嵌入。你可以在 Azure 的文档中找到他们的文档 [这里](https://docs.llamaindex.ai/en/stable/api_reference/embeddings/azure_openai/)
### Azure openai 配置
Azure OpenAI 嵌入使用的模型是 "text-embedding-ada-002"。
你需要在你的 Azure 账户中为这个模型设置正确的部署名称。
在 Docker 中运行 OpenDevin 时,使用 `-e` 设置以下环境变量:
```
LLM_EMBEDDING_MODEL="azureopenai"
LLM_EMBEDDING_DEPLOYMENT_NAME = "<your-embedding-deployment-name>" # 示例: "TextEmbedding...<etc>"
LLM_API_VERSION = "<api-version>" # 示例: "2024-02-15-preview"
```
@@ -0,0 +1,28 @@
# Google Gemini/Vertex LLM
## Completion
OpenDevin 使用 LiteLLM 进行补全调用。以下资源与使用 OpenDevin 和 Google 的 LLM 相关:
- [Gemini - Google AI Studio](https://docs.litellm.ai/docs/providers/gemini)
- [VertexAI - Google Cloud Platform](https://docs.litellm.ai/docs/providers/vertex)
### Gemini - Google AI Studio 配置
在运行 OpenDevin Docker 镜像时,通过 Google AI Studio 使用 Gemini,你需要使用 `-e` 设置以下环境变量:
```
GEMINI_API_KEY="<your-google-api-key>"
LLM_MODEL="gemini/gemini-1.5-pro"
```
### Vertex AI - Google Cloud Platform 配置
在运行 OpenDevin Docker 镜像时,通过 Google Cloud Platform 使用 Vertex AI,你需要使用 `-e` 设置以下环境变量:
```
GOOGLE_APPLICATION_CREDENTIALS="<json-dump-of-gcp-service-account-json>"
VERTEXAI_PROJECT="<your-gcp-project-id>"
VERTEXAI_LOCATION="<your-gcp-location>"
LLM_MODEL="vertex_ai/<desired-llm-model>"
```
@@ -0,0 +1,46 @@
---
sidebar_position: 2
---
# 🤖 LLM 支持
OpenDevin 可以兼容任何 LLM 后端。
关于所有可用 LM 提供商和模型的完整列表,请参阅
[litellm 文档](https://docs.litellm.ai/docs/providers)。
:::warning
OpenDevin 将向你配置的 LLM 发出许多提示。大多数这些 LLM 都是收费的——请务必设定支出限额并监控使用情况。
:::
`LLM_MODEL` 环境变量控制在编程交互中使用的模型。
但在使用 OpenDevin UI 时,你需要在设置窗口中选择你的模型(左下角的齿轮)。
某些 LLM 可能需要以下环境变量:
- `LLM_API_KEY`
- `LLM_BASE_URL`
- `LLM_EMBEDDING_MODEL`
- `LLM_EMBEDDING_DEPLOYMENT_NAME`
- `LLM_API_VERSION`
我们有一些指南,介绍了如何使用特定模型提供商运行 OpenDevin:
- [ollama](llms/localLLMs)
- [Azure](llms/azureLLMs)
如果你使用其他提供商,我们鼓励你打开一个 PR 来分享你的配置!
## 关于替代模型的注意事项
最好的模型是 GPT-4 和 Claude 3。目前的本地和开源模型
远没有那么强大。当使用替代模型时,
你可能会看到信息之间的长时间等待,
糟糕的响应,或关于 JSON格式错误的错误。OpenDevin
的强大程度依赖于其驱动的模型——幸运的是,我们团队的人员
正在积极致力于构建更好的开源模型!
## API 重试和速率限制
一些 LLM 有速率限制,可能需要重试操作。OpenDevin 会在收到 429 错误或 API 连接错误时自动重试请求。
你可以设置 `LLM_NUM_RETRIES``LLM_RETRY_MIN_WAIT``LLM_RETRY_MAX_WAIT` 环境变量来控制重试次数和重试之间的时间。
默认情况下,`LLM_NUM_RETRIES` 为 5`LLM_RETRY_MIN_WAIT``LLM_RETRY_MAX_WAIT` 分别为 3 秒和 60 秒。
@@ -0,0 +1,140 @@
# 使用 Ollama 的本地 LLM
确保您的 Ollama 服务器已启动运行。有关详细的启动说明,请参阅[此处](https://github.com/ollama/ollama)
本指南假定您已通过 `ollama serve` 启动 ollama。如果您以其他方式运行 ollama(例如在 docker 内),说明可能需要进行修改。请注意,如果您在运行 WSL,默认的 ollama 配置会阻止来自 docker 容器的请求。请参阅[此处](#configuring-ollama-service-zh-Hans)。
## 拉取模型
Ollama 模型名称可以在[这里](https://ollama.com/library)找到。一个小例子,您可以使用
`codellama:7b` 模型。较大的模型通常表现更好。
```bash
ollama pull codellama:7b
```
您可以这样检查已下载的模型:
```bash
~$ ollama list
NAME ID SIZE MODIFIED
codellama:7b 8fdf8f752f6e 3.8 GB 6 weeks ago
mistral:7b-instruct-v0.2-q4_K_M eb14864c7427 4.4 GB 2 weeks ago
starcoder2:latest f67ae0f64584 1.7 GB 19 hours ago
```
## 启动 OpenDevin
### Docker
使用[此处](../intro)的说明通过 Docker 启动 OpenDevin。
但是在运行 `docker run` 时,您需要添加一些额外的参数:
```bash
--add-host host.docker.internal:host-gateway \
-e LLM_API_KEY="ollama" \
-e LLM_BASE_URL="http://host.docker.internal:11434" \
```
例如:
```bash
# 您希望 OpenDevin 修改的目录。必须是绝对路径!
export WORKSPACE_BASE=$(pwd)/workspace
docker run \
-it \
--pull=always \
--add-host host.docker.internal:host-gateway \
-e SANDBOX_USER_ID=$(id -u) \
-e LLM_API_KEY="ollama" \
-e LLM_BASE_URL="http://host.docker.internal:11434" \
-e WORKSPACE_MOUNT_PATH=$WORKSPACE_BASE \
-v $WORKSPACE_BASE:/opt/workspace_base \
-v /var/run/docker.sock:/var/run/docker.sock \
-p 3000:3000 \
ghcr.io/opendevin/opendevin:main
```
现在您应该可以连接到 `http://localhost:3000/`
### 从源代码构建
使用[Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md)中的说明构建 OpenDevin。
通过运行 `make setup-config` 确保 `config.toml` 存在,这将为您创建一个。在 `config.toml` 中,输入以下内容:
```
LLM_MODEL="ollama/codellama:7b"
LLM_API_KEY="ollama"
LLM_EMBEDDING_MODEL="local"
LLM_BASE_URL="http://localhost:11434"
WORKSPACE_BASE="./workspace"
WORKSPACE_DIR="$(pwd)/workspace"
```
如有需要,可以替换您选择的 `LLM_MODEL`
完成!现在您可以通过 `make run` 启动 Devin 而无需 Docker。现在您应该可以连接到 `http://localhost:3000/`
## 选择您的模型
在 OpenDevin UI 中,点击左下角的设置齿轮。
然后在 `Model` 输入中,输入 `ollama/codellama:7b`,或者您之前拉取的模型名称。
如果它没有出现在下拉列表中,也没关系,只需输入即可。完成后点击保存。
现在您已经准备好了!
## 配置 ollama 服务(WSL{#configuring-ollama-service-zh-Hans}
WSL 中 ollama 的默认配置仅为 localhost 提供服务。这意味着您无法从 docker 容器中访问它。比如,它不会与 OpenDevin 一起工作。首先让我们测试 ollama 是否正常运行。
```bash
ollama list # 获取已安装模型列表
curl http://localhost:11434/api/generate -d '{"model":"[NAME]","prompt":"hi"}'
#例如,curl http://localhost:11434/api/generate -d '{"model":"codellama:7b","prompt":"hi"}'
#例如,curl http://localhost:11434/api/generate -d '{"model":"codellama","prompt":"hi"}' # 如果只有一个模型,标签是可选的
```
完成后,测试它是否允许“外部”请求,比如那些来自 docker 容器内的请求。
```bash
docker ps # 获取正在运行的 docker 容器列表,最准确的测试选择 open devin sandbox 容器。
docker exec [CONTAINER ID] curl http://host.docker.internal:11434/api/generate -d '{"model":"[NAME]","prompt":"hi"}'
#例如,docker exec cd9cc82f7a11 curl http://host.docker.internal:11434/api/generate -d '{"model":"codellama","prompt":"hi"}'
```
## 修复它
现在让我们使其工作。使用 sudo 权限编辑 /etc/systemd/system/ollama.service。 (路径可能因 linux 版本而异)
```bash
sudo vi /etc/systemd/system/ollama.service
```
或者
```bash
sudo nano /etc/systemd/system/ollama.service
```
在 [Service] 括号内添加以下行
```
Environment="OLLAMA_HOST=0.0.0.0:11434"
Environment="OLLAMA_ORIGINS=*"
```
然后保存,重新加载配置并重新启动服务。
```bash
sudo systemctl daemon-reload
sudo systemctl restart ollama
```
最后测试 ollama 是否可以从容器内访问
```bash
ollama list # 获取已安装模型列表
docker ps # 获取正在运行的 docker 容器列表,最准确的测试选择 open devin sandbox 容器。
docker exec [CONTAINER ID] curl http://host.docker.internal:11434/api/generate -d '{"model":"[NAME]","prompt":"hi"}'
```
@@ -0,0 +1,193 @@
---
sidebar_position: 5
---
# 🚧 故障排除
以下是用户经常报告的一些错误信息。
我们将努力使安装过程更加简单,并改善这些错误信息。不过,现在您可以在下面找到您的错误信息,并查看是否有任何解决方法。
对于这些错误信息,**都已经有相关的报告**。请不要打开新的报告——只需在现有的报告中发表评论即可。
如果您发现更多信息或者一个解决方法,请提交一个 *PR* 来添加细节到这个文件中。
:::tip
如果您在 Windows 上运行并遇到问题,请查看我们的[Windows (WSL) 用户指南](troubleshooting/windows)。
:::
## 无法连接到 Docker
[GitHub 问题](https://github.com/OpenDevin/OpenDevin/issues/1226)
### 症状
```bash
创建控制器时出错。请检查 Docker 是否正在运行,并访问 `https://opendevin.github.io/OpenDevin/modules/usage/troubleshooting` 获取更多调试信息。
```
```bash
docker.errors.DockerException: 获取服务器 API 版本时出错: ('连接中止。', FileNotFoundError(2, '没有这样的文件或目录'))
```
### 详情
OpenDevin 使用 Docker 容器来安全地完成工作,而不会破坏您的机器。
### 解决方法
* 运行 `docker ps` 以确保 Docker 正在运行
* 确保您不需要使用 `sudo` 运行 Docker [请参见此处](https://www.baeldung.com/linux/docker-run-without-sudo)
* 如果您使用的是 Mac,请检查[权限要求](https://docs.docker.com/desktop/mac/permission-requirements/) ,特别是考虑在 Docker Desktop 的 `Settings > Advanced` 下启用 `Allow the default Docker socket to be used`
* 另外,升级您的 Docker 到最新版本,选择 `Check for Updates`
## 无法连接到 DockerSSHBox
[GitHub 问题](https://github.com/OpenDevin/OpenDevin/issues/1156)
### 症状
```python
self.shell = DockerSSHBox(
...
pexpect.pxssh.ExceptionPxssh: Could not establish connection to host
```
### 详情
默认情况下,OpenDevin 使用 SSH 连接到一个运行中的容器。在某些机器上,尤其是 Windows,这似乎会失败。
### 解决方法
* 重新启动您的计算机(有时会有用)
* 确保拥有最新版本的 WSL 和 Docker
* 检查您的 WSL 分发版也已更新
* 尝试[此重新安装指南](https://github.com/OpenDevin/OpenDevin/issues/1156#issuecomment-2064549427)
## 无法连接到 LLM
[GitHub 问题](https://github.com/OpenDevin/OpenDevin/issues/1208)
### 症状
```python
File "/app/.venv/lib/python3.12/site-packages/openai/_exceptions.py", line 81, in __init__
super().__init__(message, response.request, body=body)
^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'request'
```
### 详情
[GitHub 问题](https://github.com/OpenDevin/OpenDevin/issues?q=is%3Aissue+is%3Aopen+404)
这通常发生在本地 LLM 设置中,当 OpenDevin 无法连接到 LLM 服务器时。请参阅我们的 [本地 LLM 指南](llms/localLLMs) 以获取更多信息。
### 解决方法
* 检查您的 `config.toml` 文件中 "llm" 部分的 `base_url` 是否正确(如果存在)
* 检查 Ollama(或您使用的其他 LLM)是否正常运行
* 确保在 Docker 中运行时使用 `--add-host host.docker.internal:host-gateway`
## `404 Resource not found 资源未找到`
### 症状
```python
Traceback (most recent call last):
File "/app/.venv/lib/python3.12/site-packages/litellm/llms/openai.py", line 414, in completion
raise e
File "/app/.venv/lib/python3.12/site-packages/litellm/llms/openai.py", line 373, in completion
response = openai_client.chat.completions.create(**data, timeout=timeout) # type: ignore
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_utils/_utils.py", line 277, in wrapper
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/resources/chat/completions.py", line 579, in create
return self._post(
^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1232, in post
return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 921, in request
return self._request(
^^^^^^^^^^^^^^
File "/app/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1012, in _request
raise self._make_status_error_from_response(err.response) from None
openai.NotFoundError: Error code: 404 - {'error': {'code': '404', 'message': 'Resource not found'}}
```
### 详情
当 LiteLLM(我们用于连接不同 LLM 提供商的库)找不到您要连接的 API 端点时,会发生这种情况。最常见的情况是 Azure 或 Ollama 用户。
### 解决方法
* 检查您是否正确设置了 `LLM_BASE_URL`
* 检查模型是否正确设置,基于 [LiteLLM 文档](https://docs.litellm.ai/docs/providers)
* 如果您在 UI 中运行,请确保在设置模式中设置 `model`
* 如果您通过 main.py 运行,请确保在环境变量/配置中设置 `LLM_MODEL`
* 确保遵循了您的 LLM 提供商的任何特殊说明
* [Ollama](/zh-Hans/modules/usage/llms/localLLMs)
* [Azure](/zh-Hans/modules/usage/llms/azureLLMs)
* [Google](/zh-Hans/modules/usage/llms/googleLLMs)
* 确保您的 API 密钥正确无误
* 尝试使用 `curl` 连接到 LLM
* 尝试[直接通过 LiteLLM 连接](https://github.com/BerriAI/litellm)来测试您的设置
## `make build` 在安装包时卡住
### 症状
安装包时卡在 `Pending...`,没有任何错误信息:
```bash
Package operations: 286 installs, 0 updates, 0 removals
- Installing certifi (2024.2.2): Pending...
- Installing h11 (0.14.0): Pending...
- Installing idna (3.7): Pending...
- Installing sniffio (1.3.1): Pending...
- Installing typing-extensions (4.11.0): Pending...
```
### 详情
在极少数情况下,`make build` 在安装包时似乎会卡住,没有任何错误信息。
### 解决方法
* 包管理器 Poetry 可能会错过用于查找凭据的配置设置(keyring)。
### 解决方法
首先使用 `env` 检查是否存在 `PYTHON_KEYRING_BACKEND` 的值。如果不存在,运行以下命令将其设置为已知值,然后重试构建:
```bash
export PYTHON_KEYRING_BACKEND=keyring.backends.null.Keyring
```
## 会话未恢复
### 症状
通常情况下,当打开 UI 时,OpenDevin 会询问是否要恢复或开始新会话。但点击“恢复”仍然会开始一个全新的聊天。
### 详情
按今天的标准安装,会话数据存储在内存中。目前,如果 OpenDevin 的服务重启,以前的会话将失效(生成一个新秘密),因此无法恢复。
### 解决方法
* 通过编辑 OpenDevin 根文件夹中的 `config.toml` 文件,更改配置以使会话持久化,指定一个 `file_store` 和一个绝对路径的 `file_store_path`
```toml
file_store="local"
file_store_path="/absolute/path/to/opendevin/cache/directory"
```
* 在您的 .bashrc 中添加一个固定的 JWT 秘密,如下所示,以便以前的会话 ID 可以被接受。
```bash
EXPORT JWT_SECRET=A_CONST_VALUE
```
@@ -0,0 +1,76 @@
# Windows 和 WSL 用户须知
OpenDevin 仅支持通过 [WSL](https://learn.microsoft.com/en-us/windows/wsl/install) 在 Windows 上运行。
请确保在 WSL 终端内运行所有命令。
## 故障排除
### 错误:在此 WSL 2 发行版中找不到 'docker'。
如果您使用的是 Docker Desktop,请确保在 WSL 内部调用任何 docker 命令之前启动它。
Docker 还需要启用 WSL 集成选项。
### 建议:不要以 root 用户身份运行
出于安全原因,非常建议不要以 root 用户身份运行 OpenDevin,而是使用 UID 非零的用户身份运行。
此外,当以 root 身份运行时,不支持持久化沙箱,并且在启动 OpenDevin 时可能会出现相应消息。
参考资料:
* [为什么以 root 登录是不好的](https://askubuntu.com/questions/16178/why-is-it-bad-to-log-in-as-root)
* [在 WSL 中设置默认用户](https://www.tenforums.com/tutorials/128152-set-default-user-windows-subsystem-linux-distro-windows-10-a.html#option2)
关于第二个参考资料的小提示:对于 Ubuntu 用户,该命令实际上可能是 "ubuntupreview" 而不是 "ubuntu"。
### 创建 opendevin 用户失败
如果您在设置过程中遇到以下错误:
```sh
Exception: Failed to create opendevin user in sandbox: 'useradd: UID 0 is not unique'
```
您可以通过运行以下命令解决:
```sh
export SANDBOX_USER_ID=1000
```
### Poetry 安装
* 如果在构建过程中安装 Poetry 后仍然面临运行 Poetry 的问题,您可能需要将其二进制路径添加到您的环境变量:
```sh
export PATH="$HOME/.local/bin:$PATH"
```
* 如果 make build 停止并出现如下错误:
```sh
ModuleNotFoundError: no module named <module-name>
```
这可能是 Poetry 缓存的问题。
尝试运行以下两个命令:
```sh
rm -r ~/.cache/pypoetry
make build
```
### NoneType 对象没有属性 'request'
如果您在执行 `make run` 时遇到与网络相关的问题,例如 `NoneType object has no attribute 'request'`,您可能需要配置您的 WSL2 网络设置。请按照以下步骤操作:
* 打开或创建位于 Windows 主机机器上的 `C:\Users\%username%\.wslconfig` 文件。
*`.wslconfig` 文件添加以下配置:
```sh
[wsl2]
networkingMode=mirrored
localhostForwarding=true
```
* 保存 `.wslconfig` 文件。
* 通过退出所有正在运行的 WSL2 实例并在命令提示符或终端中执行 `wsl --shutdown` 命令,完全重启 WSL2。
* 重新启动 WSL 后,尝试再次执行 `make run`
网络问题应该已经解决。
@@ -0,0 +1,26 @@
{
"title": {
"message": "OpenDevin",
"description": "The title in the navbar"
},
"logo.alt": {
"message": "OpenDevin",
"description": "The alt text of navbar logo"
},
"item.label.Docs": {
"message": "文档",
"description": "Navbar item with label Docs"
},
"item.label.Codebase": {
"message": "代码库",
"description": "Navbar item with label Codebase"
},
"item.label.FAQ": {
"message": "常见问题",
"description": "Navbar item with label FAQ"
},
"item.label.GitHub": {
"message": "GitHub",
"description": "Navbar item with label GitHub"
}
}
+2 -2
View File
@@ -1,5 +1,5 @@
---
sidebar_position: 7
sidebar_position: 8
---
# 📚 Misc
@@ -29,7 +29,7 @@ For details, please check [this document](https://github.com/OpenDevin/OpenDevin
## 🤖 Join Our Community
Now we have both Slack workspace for the collaboration on building OpenDevin and Discord server for discussion about anything related, e.g., this project, LLM, agent, etc.
We have both Slack workspace for the collaboration on building OpenDevin and Discord server for discussion about anything related, e.g., this project, LLM, agent, etc.
- [Slack workspace](https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA)
- [Discord server](https://discord.gg/ESHStjSjD4)
-13
View File
@@ -55,12 +55,6 @@ _Example of CodeActAgent with `gpt-4-turbo-2024-04-09` performing a data science
| --------------- | ----------------------------------------------------------------------------------------------------------------------------------------------- |
| `__init__` | Initializes an agent with `llm` and a list of messages `list[Mapping[str, str]]` |
| `step` | Performs one step using the CodeAct Agent. This includes gathering info on previous steps and prompting the model to make a command to execute. |
| `search_memory` | Not yet implemented |
### Work-in-progress &amp; Next step
[] Support web-browsing
[] Complete the workflow for CodeAct agent to submit Github PRs
## Monologue Agent
@@ -77,7 +71,6 @@ Short term memory is stored as a Monologue object and the model can condense it
`CmdRunAction`,
`FileWriteAction`,
`FileReadAction`,
`AgentRecallAction`,
`BrowseURLAction`,
`GithubPushAction`,
`AgentThinkAction`
@@ -88,7 +81,6 @@ Short term memory is stored as a Monologue object and the model can condense it
`NullObservation`,
`CmdOutputObservation`,
`FileReadObservation`,
`AgentRecallObservation`,
`BrowserOutputObservation`
### Methods
@@ -99,7 +91,6 @@ Short term memory is stored as a Monologue object and the model can condense it
| `_add_event` | Appends events to the monologue of the agent and condenses with summary automatically if the monologue is too long |
| `_initialize` | Utilizes the `INITIAL_THOUGHTS` list to give the agent a context for its capabilities and how to navigate the `/workspace` |
| `step` | Modifies the current state by adding the most recent actions and observations, then prompts the model to think about its next action to take. |
| `search_memory` | Uses `VectorIndexRetriever` to find related memories within the long term memory. |
## Planner Agent
@@ -112,12 +103,10 @@ The agent is given its previous action-observation pairs, current task, and hint
`NullAction`,
`CmdRunAction`,
`CmdKillAction`,
`BrowseURLAction`,
`GithubPushAction`,
`FileReadAction`,
`FileWriteAction`,
`AgentRecallAction`,
`AgentThinkAction`,
`AgentFinishAction`,
`AgentSummarizeAction`,
@@ -130,7 +119,6 @@ The agent is given its previous action-observation pairs, current task, and hint
`NullObservation`,
`CmdOutputObservation`,
`FileReadObservation`,
`AgentRecallObservation`,
`BrowserOutputObservation`
### Methods
@@ -139,4 +127,3 @@ The agent is given its previous action-observation pairs, current task, and hint
| --------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `__init__` | Initializes an agent with `llm` |
| `step` | Checks to see if current step is completed, returns `AgentFinishAction` if True. Otherwise, creates a plan prompt and sends to model for inference, adding the result as the next action. |
| `search_memory` | Not yet implemented |
+10 -3
View File
@@ -1,16 +1,23 @@
---
sidebar_position: 4
sidebar_position: 7
---
# 🏛️ System Architecture Overview
<div style={{ textAlign: 'center' }}>
<img src="https://github.com/OpenDevin/OpenDevin/assets/16201837/97d747e3-29d8-4ccb-8d34-6ad1adb17f38" alt="OpenDevin System Architecture Diagram Jul 4 2024" />
<p><em>OpenDevin System Architecture Diagram (July 4, 2024)</em></p>
</div>
This is a high-level overview of the system architecture. The system is divided into two main components: the frontend and the backend. The frontend is responsible for handling user interactions and displaying the results. The backend is responsible for handling the business logic and executing the agents.
# Frontend Architecture {#frontend-architecture-en}
![system_architecture.svg](/img/system_architecture.svg)
This Overview is simplified to show the main components and their interactions. For a more detailed view of the backend architecture, see the [Backend Architecture](#backend-architecture) section.
This Overview is simplified to show the main components and their interactions. For a more detailed view of the backend architecture, see the Backend Architecture section below.
# Backend Architecture
# Backend Architecture {#backend-architecture-en}
_**Disclaimer**: The backend architecture is a work in progress and is subject to change. The following diagram shows the current architecture of the backend based on the commit that is shown in the footer of the diagram._
+75
View File
@@ -0,0 +1,75 @@
---
sidebar_position: 8
---
# Changelog
## 0.8 (release date: ??)
### Config breaking changes
In this release we introduced a few breaking changes to backend configurations.
If you have only been using OpenDevin via frontend (web GUI), nothing needs
to be taken care of.
Here's a list of breaking changes in configs. They only apply to users who
use OpenDevin CLI via `main.py`. For more detail, see [#2756](https://github.com/OpenDevin/OpenDevin/pull/2756).
#### Removal of --model-name option from main.py
Please note that `--model-name`, or `-m` option, no longer exists. You should set up the LLM
configs in `config.toml` or via environmental variables.
#### LLM config groups must be subgroups of 'llm'
Prior to release 0.8, you can use arbitrary name for llm config in `config.toml`, e.g.
```toml
[gpt-4o]
model="gpt-4o"
api_key="<your_api_key>"
```
and then use `--llm-config` CLI argument to specify the desired LLM config group
by name. This no longer works. Instead, the config group must be under `llm` group,
e.g.:
```toml
[llm.gpt-4o]
model="gpt-4o"
api_key="<your_api_key>"
```
If you have a config group named `llm`, no need to change it, it will be used
as the default LLM config group.
#### 'agent' group no longer contains 'name' field
Prior to release 0.8, you may or may not have a config group named `agent` that
looks like this:
```toml
[agent]
name="CodeActAgent"
memory_max_threads=2
```
Note the `name` field is now removed. Instead, you should put `default_agent` field
under `core` group, e.g.
```toml
[core]
# other configs
default_agent='CodeActAgent'
[agent]
llm_config='llm'
memory_max_threads=2
[agent.CodeActAgent]
llm_config='gpt-4o'
```
Note that similar to `llm` subgroups, you can also define `agent` subgroups.
Moreover, an agent can be associated with a specific LLM config group. For more
detail, see the examples in `config.template.toml`.
+129
View File
@@ -0,0 +1,129 @@
---
sidebar_position: 6
---
# 💿 How to Create and Use a Custom Docker Sandbox
The default OpenDevin sandbox comes with a [minimal ubuntu configuration](https://github.com/OpenDevin/OpenDevin/blob/main/containers/sandbox/Dockerfile).
Your use case may need additional software installed by default.
There are two ways you can do so:
1. Use an existing image from docker hub. For instance, if you want to have `nodejs` installed, you can do so by using the `node:20` image
2. Creating your own custom docker image and using it
If you want to take the first approach, you can skip the `Create Your Docker Image` section.
## Setup
Make sure you are able to run OpenDevin using the [Development.md](https://github.com/OpenDevin/OpenDevin/blob/main/Development.md) first.
## Create Your Docker Image
To create a custom docker image, it must be debian/ubuntu based.
For example, if we want OpenDevin to have access to the `node` binary, we would use the following Dockerfile:
```dockerfile
# Start with latest ubuntu image
FROM ubuntu:latest
# Run needed updates
RUN apt-get update && apt-get install -y
# Install node
RUN apt-get install -y nodejs
```
Next build your docker image with the name of your choice, for example `custom_image`.
To do this you can create a directory and put your file inside it with the name `Dockerfile`, and inside the directory run the following command:
```bash
docker build -t custom_image .
```
This will produce a new image called ```custom_image``` that will be available in Docker Engine.
> Note that in the configuration described in this document, OpenDevin will run as user "opendevin" inside the sandbox and thus all packages installed via the docker file should be available to all users on the system, not just root.
>
> Installing with apt-get above installs node for all users.
## Specify your sandbox image in config.toml file
OpenDevin configuration occurs via the top-level `config.toml` file.
Create a `config.toml` file in the OpenDevin directory and enter these contents:
```toml
[core]
workspace_base="./workspace"
persist_sandbox=false
run_as_devin=true
sandbox_container_image="custom_image"
```
For `sandbox_container_image`, you can specify either:
1. The name of your custom image that you built in the previous step (e.g., `”custom_image”`)
2. A pre-existing image from Docker Hub (e.g., `”node:20”` if you want a sandbox with Node.js pre-installed)
## Run
Run OpenDevin by running ```make run``` in the top level directory.
Navigate to ```localhost:3001``` and check if your desired dependencies are available.
In the case of the example above, running ```node -v``` in the terminal produces ```v20.15.0```
Congratulations!
## Technical Explanation
The relevant code is defined in [ssh_box.py](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/ssh_box.py) and [image_agnostic_util.py](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py).
In particular, ssh_box.py checks the config object for ```config.sandbox_container_image``` and then attempts to retrieve the image using [get_od_sandbox_image](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L72) which is defined in image_agnostic_util.py.
When first using a custom image, it will not be found and thus it will be built (on subsequent runs the built image will be found and returned).
The custom image is built using [_build_sandbox_image()](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L29), which creates a docker file using your custom_image as a base and then configures the environment for OpenDevin, like this:
```python
dockerfile_content = (
f'FROM {base_image}\n'
'RUN apt update && apt install -y openssh-server wget sudo\n'
'RUN mkdir -p -m0755 /var/run/sshd\n'
'RUN mkdir -p /opendevin && mkdir -p /opendevin/logs && chmod 777 /opendevin/logs\n'
'RUN echo "" > /opendevin/bash.bashrc\n'
'RUN if [ ! -d /opendevin/miniforge3 ]; then \\\n'
' wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh" && \\\n'
' bash Miniforge3-$(uname)-$(uname -m).sh -b -p /opendevin/miniforge3 && \\\n'
' chmod -R g+w /opendevin/miniforge3 && \\\n'
' bash -c ". /opendevin/miniforge3/etc/profile.d/conda.sh && conda config --set changeps1 False && conda config --append channels conda-forge"; \\\n'
' fi\n'
'RUN /opendevin/miniforge3/bin/pip install --upgrade pip\n'
'RUN /opendevin/miniforge3/bin/pip install jupyterlab notebook jupyter_kernel_gateway flake8\n'
'RUN /opendevin/miniforge3/bin/pip install python-docx PyPDF2 python-pptx pylatexenc openai\n'
).strip()
```
> Note: the name of the image is modified via [_get_new_image_name()](https://github.com/OpenDevin/OpenDevin/blob/main/opendevin/runtime/docker/image_agnostic_util.py#L63) and it is the modified name that is searched for on subsequent runs.
## Troubleshooting / Errors
### Error: ```useradd: UID 1000 is not unique```
If you see this error in the console output it is because OpenDevin is trying to create the opendevin user in the sandbox with a UID of 1000, however this UID is already being used in the image (for some reason). To fix this change the sandbox_user_id field in the config.toml file to a different value:
```toml
[core]
workspace_base="./workspace"
persist_sandbox=false
run_as_devin=true
sandbox_container_image="custom_image"
sandbox_user_id="1001"
```
### Port use errors
If you see an error about a port being in use or unavailable, try deleting all running Docker Containers (run `docker ps` and `docker rm` relevant containers) and then re-running ```make run```
## Discuss
For other issues or questions join the [Slack](https://join.slack.com/t/opendevin/shared_invite/zt-2jsrl32uf-fTeeFjNyNYxqSZt5NPY3fA) or [Discord](https://discord.gg/ESHStjSjD4) and ask!

Some files were not shown because too many files have changed in this diff Show More