Compare commits

..

1 Commits

Author SHA1 Message Date
Luke
ec47a318bc Release v0.4.6 (#5039) 2023-07-28 14:37:32 +02:00
2993 changed files with 20357 additions and 890137 deletions

View File

@@ -1,2 +1,2 @@
[run]
[run]
relative_files = true

View File

@@ -10,4 +10,4 @@ RUN apt-get update && apt-get install -y \
RUN apt-get install -y curl jq wget git
# Declare working directory
WORKDIR /workspace/AutoGPT
WORKDIR /workspace/Auto-GPT

View File

@@ -1,7 +1,7 @@
{
"dockerComposeFile": "./docker-compose.yml",
"service": "auto-gpt",
"workspaceFolder": "/workspace/AutoGPT",
"workspaceFolder": "/workspace/Auto-GPT",
"shutdownAction": "stopCompose",
"features": {
"ghcr.io/devcontainers/features/common-utils:2": {
@@ -46,11 +46,11 @@
// "forwardPorts": [],
// Use 'postCreateCommand' to run commands after the container is created.
// "postCreateCommand": "poetry install",
// "postCreateCommand": "pip3 install --user -r requirements.txt",
// Set `remoteUser` to `root` to connect as root instead. More info: https://aka.ms/vscode-remote/containers/non-root.
"remoteUser": "vscode",
// Add the freshly containerized repo to the list of safe repositories
"postCreateCommand": "git config --global --add safe.directory /workspace/AutoGPT && poetry install"
}
"postCreateCommand": "git config --global --add safe.directory /workspace/Auto-GPT && pip3 install --user -r requirements.txt"
}

View File

@@ -9,4 +9,4 @@ services:
context: ../
tty: true
volumes:
- ../:/workspace/AutoGPT
- ../:/workspace/Auto-GPT

9
.dockerignore Normal file
View File

@@ -0,0 +1,9 @@
.*
*.template
*.yaml
*.yml
!prompt_settings.yaml
*.md
*.png
!BULLETIN.md

212
.env.template Normal file
View File

@@ -0,0 +1,212 @@
# For further descriptions of these settings see docs/configuration/options.md or go to docs.agpt.co
################################################################################
### AUTO-GPT - GENERAL SETTINGS
################################################################################
## OPENAI_API_KEY - OpenAI API Key (Example: my-openai-api-key)
OPENAI_API_KEY=your-openai-api-key
## EXECUTE_LOCAL_COMMANDS - Allow local command execution (Default: False)
# EXECUTE_LOCAL_COMMANDS=False
## RESTRICT_TO_WORKSPACE - Restrict file operations to workspace ./auto_gpt_workspace (Default: True)
# RESTRICT_TO_WORKSPACE=True
## USER_AGENT - Define the user-agent used by the requests library to browse website (string)
# USER_AGENT="Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36"
## AI_SETTINGS_FILE - Specifies which AI Settings file to use, relative to the Auto-GPT root directory. (defaults to ai_settings.yaml)
# AI_SETTINGS_FILE=ai_settings.yaml
## PLUGINS_CONFIG_FILE - The path to the plugins_config.yaml file, relative to the Auto-GPT root directory. (Default plugins_config.yaml)
# PLUGINS_CONFIG_FILE=plugins_config.yaml
## PROMPT_SETTINGS_FILE - Specifies which Prompt Settings file to use, relative to the Auto-GPT root directory. (defaults to prompt_settings.yaml)
# PROMPT_SETTINGS_FILE=prompt_settings.yaml
## OPENAI_API_BASE_URL - Custom url for the OpenAI API, useful for connecting to custom backends. No effect if USE_AZURE is true, leave blank to keep the default url
# the following is an example:
# OPENAI_API_BASE_URL=http://localhost:443/v1
## OPENAI_FUNCTIONS - Enables OpenAI functions: https://platform.openai.com/docs/guides/gpt/function-calling
## WARNING: this feature is only supported by OpenAI's newest models. Until these models become the default on 27 June, add a '-0613' suffix to the model of your choosing.
# OPENAI_FUNCTIONS=False
## AUTHORISE COMMAND KEY - Key to authorise commands
# AUTHORISE_COMMAND_KEY=y
## EXIT_KEY - Key to exit AUTO-GPT
# EXIT_KEY=n
## PLAIN_OUTPUT - Plain output, which disables the spinner (Default: False)
# PLAIN_OUTPUT=False
## DISABLED_COMMAND_CATEGORIES - The list of categories of commands that are disabled (Default: None)
# DISABLED_COMMAND_CATEGORIES=
################################################################################
### LLM PROVIDER
################################################################################
## TEMPERATURE - Sets temperature in OpenAI (Default: 0)
# TEMPERATURE=0
## OPENAI_ORGANIZATION - Your OpenAI Organization key (Default: None)
# OPENAI_ORGANIZATION=
## USE_AZURE - Use Azure OpenAI or not (Default: False)
# USE_AZURE=False
## AZURE_CONFIG_FILE - The path to the azure.yaml file, relative to the Auto-GPT root directory. (Default: azure.yaml)
# AZURE_CONFIG_FILE=azure.yaml
################################################################################
### LLM MODELS
################################################################################
## SMART_LLM - Smart language model (Default: gpt-4)
# SMART_LLM=gpt-4
## FAST_LLM - Fast language model (Default: gpt-3.5-turbo)
# FAST_LLM=gpt-3.5-turbo
## EMBEDDING_MODEL - Model to use for creating embeddings
# EMBEDDING_MODEL=text-embedding-ada-002
################################################################################
### SHELL EXECUTION
################################################################################
## SHELL_COMMAND_CONTROL - Whether to use "allowlist" or "denylist" to determine what shell commands can be executed (Default: denylist)
# SHELL_COMMAND_CONTROL=denylist
## ONLY if SHELL_COMMAND_CONTROL is set to denylist:
## SHELL_DENYLIST - List of shell commands that ARE NOT allowed to be executed by Auto-GPT (Default: sudo,su)
# SHELL_DENYLIST=sudo,su
## ONLY if SHELL_COMMAND_CONTROL is set to allowlist:
## SHELL_ALLOWLIST - List of shell commands that ARE allowed to be executed by Auto-GPT (Default: None)
# SHELL_ALLOWLIST=
################################################################################
### MEMORY
################################################################################
### General
## MEMORY_BACKEND - Memory backend type
# MEMORY_BACKEND=json_file
## MEMORY_INDEX - Value used in the Memory backend for scoping, naming, or indexing (Default: auto-gpt)
# MEMORY_INDEX=auto-gpt
### Redis
## REDIS_HOST - Redis host (Default: localhost, use "redis" for docker-compose)
# REDIS_HOST=localhost
## REDIS_PORT - Redis port (Default: 6379)
# REDIS_PORT=6379
## REDIS_PASSWORD - Redis password (Default: "")
# REDIS_PASSWORD=
## WIPE_REDIS_ON_START - Wipes data / index on start (Default: True)
# WIPE_REDIS_ON_START=True
################################################################################
### IMAGE GENERATION PROVIDER
################################################################################
### Common
## IMAGE_PROVIDER - Image provider (Default: dalle)
# IMAGE_PROVIDER=dalle
## IMAGE_SIZE - Image size (Default: 256)
# IMAGE_SIZE=256
### Huggingface (IMAGE_PROVIDER=huggingface)
## HUGGINGFACE_IMAGE_MODEL - Text-to-image model from Huggingface (Default: CompVis/stable-diffusion-v1-4)
# HUGGINGFACE_IMAGE_MODEL=CompVis/stable-diffusion-v1-4
## HUGGINGFACE_API_TOKEN - HuggingFace API token (Default: None)
# HUGGINGFACE_API_TOKEN=
### Stable Diffusion (IMAGE_PROVIDER=sdwebui)
## SD_WEBUI_AUTH - Stable Diffusion Web UI username:password pair (Default: None)
# SD_WEBUI_AUTH=
## SD_WEBUI_URL - Stable Diffusion Web UI API URL (Default: http://localhost:7860)
# SD_WEBUI_URL=http://localhost:7860
################################################################################
### AUDIO TO TEXT PROVIDER
################################################################################
## AUDIO_TO_TEXT_PROVIDER - Audio-to-text provider (Default: huggingface)
# AUDIO_TO_TEXT_PROVIDER=huggingface
## HUGGINGFACE_AUDIO_TO_TEXT_MODEL - The model for HuggingFace to use (Default: CompVis/stable-diffusion-v1-4)
# HUGGINGFACE_AUDIO_TO_TEXT_MODEL=CompVis/stable-diffusion-v1-4
################################################################################
### GITHUB
################################################################################
## GITHUB_API_KEY - Github API key / PAT (Default: None)
# GITHUB_API_KEY=
## GITHUB_USERNAME - Github username (Default: None)
# GITHUB_USERNAME=
################################################################################
### WEB BROWSING
################################################################################
## HEADLESS_BROWSER - Whether to run the browser in headless mode (default: True)
# HEADLESS_BROWSER=True
## USE_WEB_BROWSER - Sets the web-browser driver to use with selenium (default: chrome)
# USE_WEB_BROWSER=chrome
## BROWSE_CHUNK_MAX_LENGTH - When browsing website, define the length of chunks to summarize (Default: 3000)
# BROWSE_CHUNK_MAX_LENGTH=3000
## BROWSE_SPACY_LANGUAGE_MODEL - spaCy language model](https://spacy.io/usage/models) to use when creating chunks. (Default: en_core_web_sm)
# BROWSE_SPACY_LANGUAGE_MODEL=en_core_web_sm
## GOOGLE_API_KEY - Google API key (Default: None)
# GOOGLE_API_KEY=
## GOOGLE_CUSTOM_SEARCH_ENGINE_ID - Google custom search engine ID (Default: None)
# GOOGLE_CUSTOM_SEARCH_ENGINE_ID=
################################################################################
### TEXT TO SPEECH PROVIDER
################################################################################
## TEXT_TO_SPEECH_PROVIDER - Which Text to Speech provider to use (Default: gtts)
# TEXT_TO_SPEECH_PROVIDER=gtts
### Only if TEXT_TO_SPEECH_PROVIDER=streamelements
## STREAMELEMENTS_VOICE - Voice to use for StreamElements (Default: Brian)
# STREAMELEMENTS_VOICE=Brian
### Only if TEXT_TO_SPEECH_PROVIDER=elevenlabs
## ELEVENLABS_API_KEY - Eleven Labs API key (Default: None)
# ELEVENLABS_API_KEY=
## ELEVENLABS_VOICE_ID - Eleven Labs voice ID (Example: None)
# ELEVENLABS_VOICE_ID=
################################################################################
### CHAT MESSAGES
################################################################################
## CHAT_MESSAGES_ENABLED - Enable chat messages (Default: False)
# CHAT_MESSAGES_ENABLED=False

View File

@@ -1,4 +1,4 @@
# Upon entering directory, direnv requests user permission once to automatically load project dependencies onwards.
# Eliminating the need of running "nix develop github:superherointj/nix-auto-gpt" for Nix users to develop/use AutoGPT.
# Eliminating the need of running "nix develop github:superherointj/nix-auto-gpt" for Nix users to develop/use Auto-GPT.
[[ -z $IN_NIX_SHELL ]] && use flake github:superherointj/nix-auto-gpt

View File

@@ -1,5 +1,6 @@
[flake8]
max-line-length = 88
select = "E303, W293, W291, W292, E305, E231, E302"
exclude =
.tox,
__pycache__,
@@ -9,4 +10,3 @@ exclude =
.venv/*,
reports/*,
dist/*,
data/*,

11
.gitattributes vendored
View File

@@ -1,10 +1,5 @@
classic/frontend/build/** linguist-generated
**/poetry.lock linguist-generated
docs/_javascript/** linguist-vendored
# Exclude VCR cassettes from stats
classic/forge/tests/vcr_cassettes/**/**.y*ml linguist-generated
tests/Auto-GPT-test-cassettes/**/**.y*ml linguist-generated
* text=auto
# Mark documentation as such
docs/**.md linguist-documentation

9
.github/CODEOWNERS vendored
View File

@@ -1,7 +1,2 @@
* @Significant-Gravitas/maintainers
.github/workflows/ @Significant-Gravitas/devops
classic/forge/ @Significant-Gravitas/forge-maintainers
classic/benchmark/ @Significant-Gravitas/benchmark-maintainers
classic/frontend/ @Significant-Gravitas/frontend-maintainers
autogpt_platform/infra @Significant-Gravitas/devops
.github/CODEOWNERS @Significant-Gravitas/admins
.github/workflows/ @Significant-Gravitas/maintainers
autogpt/core @collijk

3
.github/FUNDING.yml vendored Normal file
View File

@@ -0,0 +1,3 @@
# These are supported funding model platforms
github: Torantulino

View File

@@ -1,5 +1,5 @@
name: Bug report 🐛
description: Create a bug report for AutoGPT.
description: Create a bug report for Auto-GPT.
labels: ['status: needs triage']
body:
- type: markdown
@@ -13,16 +13,16 @@ body:
[backlog]: https://github.com/orgs/Significant-Gravitas/projects/1
[roadmap]: https://github.com/orgs/Significant-Gravitas/projects/2
[discord]: https://discord.gg/autogpt
[discussions]: https://github.com/Significant-Gravitas/AutoGPT/discussions
[discussions]: https://github.com/Significant-Gravitas/Auto-GPT/discussions
[#tech-support]: https://discord.com/channels/1092243196446249134/1092275629602394184
[existing issues]: https://github.com/Significant-Gravitas/AutoGPT/issues?q=is%3Aissue
[wiki page on Contributing]: https://github.com/Significant-Gravitas/AutoGPT/wiki/Contributing
[existing issues]: https://github.com/Significant-Gravitas/Auto-GPT/issues?q=is%3Aissue
[wiki page on Contributing]: https://github.com/Significant-Gravitas/Nexus/wiki/Contributing
- type: checkboxes
attributes:
label: ⚠️ Search for existing issues first ⚠️
description: >
Please [search the history](https://github.com/Significant-Gravitas/AutoGPT/issues)
Please [search the history](https://github.com/Torantulino/Auto-GPT/issues)
to see if an issue already exists for the same problem.
options:
- label: I have searched the existing issues, and there is no existing issue for my problem
@@ -35,8 +35,8 @@ body:
A good rule of thumb: What would you type if you were searching for the issue?
For example:
BAD - my AutoGPT keeps looping
GOOD - After performing execute_python_file, AutoGPT goes into a loop where it keeps trying to execute the file.
BAD - my auto-gpt keeps looping
GOOD - After performing execute_python_file, auto-gpt goes into a loop where it keeps trying to execute the file.
⚠️ SUPER-busy repo, please help the volunteer maintainers.
The less time we spend here, the more time we can spend building AutoGPT.
@@ -54,7 +54,7 @@ body:
attributes:
label: Which Operating System are you using?
description: >
Please select the operating system you were using to run AutoGPT when this problem occurred.
Please select the operating system you were using to run Auto-GPT when this problem occurred.
options:
- Windows
- Linux
@@ -73,12 +73,12 @@ body:
- type: dropdown
attributes:
label: Which version of AutoGPT are you using?
label: Which version of Auto-GPT are you using?
description: |
Please select which version of AutoGPT you were using when this issue occurred.
If you downloaded the code from the [releases page](https://github.com/Significant-Gravitas/AutoGPT/releases/) make sure you were using the latest code.
**If you weren't please try with the [latest code](https://github.com/Significant-Gravitas/AutoGPT/releases/)**.
If installed with git you can run `git branch` to see which version of AutoGPT you are running.
Please select which version of Auto-GPT you were using when this issue occurred.
If you downloaded the code from the [releases page](https://github.com/Significant-Gravitas/Auto-GPT/releases/) make sure you were using the latest code.
**If you weren't please try with the [latest code](https://github.com/Significant-Gravitas/Auto-GPT/releases/)**.
If installed with git you can run `git branch` to see which version of Auto-GPT you are running.
options:
- Latest Release
- Stable (branch)
@@ -88,16 +88,14 @@ body:
- type: dropdown
attributes:
label: What LLM Provider do you use?
label: Do you use OpenAI GPT-3 or GPT-4?
description: >
If you are using AutoGPT with `SMART_LLM=gpt-3.5-turbo`, your problems may be caused by
the [limitations](https://github.com/Significant-Gravitas/AutoGPT/issues?q=is%3Aissue+label%3A%22AI+model+limitation%22) of GPT-3.5.
If you are using Auto-GPT with `--gpt3only`, your problems may be caused by
the [limitations](https://github.com/Significant-Gravitas/Auto-GPT/issues?q=is%3Aissue+label%3A%22AI+model+limitation%22) of GPT-3.5.
options:
- Azure
- Groq
- Anthropic
- Llamafile
- Other (detail in issue)
- GPT-3.5
- GPT-4
- GPT-4(32k)
validations:
required: true
@@ -128,17 +126,10 @@ body:
label: Specify the area
description: Please specify the area you think is best related to the issue.
- type: input
attributes:
label: What commit or version are you using?
description: It is helpful for us to reproduce to know what version of the software you were using when this happened. Please run `git log -n 1 --pretty=format:"%H"` to output the full commit hash.
validations:
required: true
- type: textarea
attributes:
label: Describe your issue.
description: Describe the problem you are experiencing. Try to describe only the issue and phrase it short but clear. ⚠️ Provide NO other data in this field
description: Describe the problem you are experiencing. Try to describe only the issue and phrase it short but clear. ⚠️ Provide NO other data in this field
validations:
required: true
@@ -148,16 +139,16 @@ body:
value: |
The following is OPTIONAL, please keep in mind that the log files may contain personal information such as credentials.⚠️
"The log files are located in the folder 'logs' inside the main AutoGPT folder."
"The log files are located in the folder 'logs' inside the main auto-gpt folder."
- type: textarea
attributes:
label: Upload Activity Log Content
description: |
Upload the activity log content, this can help us understand the issue better.
To do this, go to the folder logs in your main AutoGPT folder, open activity.log and copy/paste the contents to this field.
⚠️ The activity log may contain personal data given to AutoGPT by you in prompt or input as well as
any personal information that AutoGPT collected out of files during last run. Do not add the activity log if you are not comfortable with sharing it. ⚠️
To do this, go to the folder logs in your main auto-gpt folder, open activity.log and copy/paste the contents to this field.
⚠️ The activity log may contain personal data given to auto-gpt by you in prompt or input as well as
any personal information that auto-gpt collected out of files during last run. Do not add the activity log if you are not comfortable with sharing it. ⚠️
validations:
required: false
@@ -166,8 +157,8 @@ body:
label: Upload Error Log Content
description: |
Upload the error log content, this will help us understand the issue better.
To do this, go to the folder logs in your main AutoGPT folder, open error.log and copy/paste the contents to this field.
⚠️ The error log may contain personal data given to AutoGPT by you in prompt or input as well as
any personal information that AutoGPT collected out of files during last run. Do not add the activity log if you are not comfortable with sharing it. ⚠️
To do this, go to the folder logs in your main auto-gpt folder, open error.log and copy/paste the contents to this field.
⚠️ The error log may contain personal data given to auto-gpt by you in prompt or input as well as
any personal information that auto-gpt collected out of files during last run. Do not add the activity log if you are not comfortable with sharing it. ⚠️
validations:
required: false

View File

@@ -1,16 +1,16 @@
name: Feature request 🚀
description: Suggest a new idea for AutoGPT!
description: Suggest a new idea for Auto-GPT!
labels: ['status: needs triage']
body:
- type: markdown
attributes:
value: |
First, check out our [wiki page on Contributing](https://github.com/Significant-Gravitas/AutoGPT/wiki/Contributing)
First, check out our [wiki page on Contributing](https://github.com/Significant-Gravitas/Nexus/wiki/Contributing)
Please provide a searchable summary of the issue in the title above ⬆️.
- type: checkboxes
attributes:
label: Duplicates
description: Please [search the history](https://github.com/Significant-Gravitas/AutoGPT/issues) to see if an issue already exists for the same problem.
description: Please [search the history](https://github.com/Torantulino/Auto-GPT/issues) to see if an issue already exists for the same problem.
options:
- label: I have searched the existing issues
required: true

View File

@@ -1,23 +1,49 @@
### Background
<!-- Clearly explain the need for these changes: -->
### Changes 🏗️
<!-- Concisely describe all of the changes made in this pull request: -->
### Testing 🔍
> [!NOTE]
Only for the new autogpt platform, currently in autogpt_platform/
<!--
Please make sure your changes have been tested and are in good working condition.
Here is a list of our critical paths, if you need some inspiration on what and how to test:
<!-- ⚠️ At the moment any non-essential commands are not being merged.
If you want to add non-essential commands to Auto-GPT, please create a plugin instead.
We are expecting to ship plugin support within the week (PR #757).
Resources:
* https://github.com/Significant-Gravitas/Auto-GPT-Plugin-Template
-->
- Create from scratch and execute an agent with at least 3 blocks
- Import an agent from file upload, and confirm it executes correctly
- Upload agent to marketplace
- Import an agent from marketplace and confirm it executes correctly
- Edit an agent from monitor, and confirm it executes correctly
<!-- 📢 Announcement
We've recently noticed an increase in pull requests focusing on combining multiple changes. While the intentions behind these PRs are appreciated, it's essential to maintain a clean and manageable git history. To ensure the quality of our repository, we kindly ask you to adhere to the following guidelines when submitting PRs:
Focus on a single, specific change.
Do not include any unrelated or "extra" modifications.
Provide clear documentation and explanations of the changes made.
Ensure diffs are limited to the intended lines — no applying preferred formatting styles or line endings (unless that's what the PR is about).
For guidance on committing only the specific lines you have changed, refer to this helpful video: https://youtu.be/8-hSNHHbiZg
Check out our [wiki page on Contributing](https://github.com/Significant-Gravitas/Nexus/wiki/Contributing)
By following these guidelines, your PRs are more likely to be merged quickly after testing, as long as they align with the project's overall direction. -->
### Background
<!-- Provide a concise overview of the rationale behind this change. Include relevant context, prior discussions, or links to related issues. Ensure that the change aligns with the project's overall direction. -->
### Changes
<!-- Describe the specific, focused change made in this pull request. Detail the modifications clearly and avoid any unrelated or "extra" changes. -->
### Documentation
<!-- Explain how your changes are documented, such as in-code comments or external documentation. Ensure that the documentation is clear, concise, and easy to understand. -->
### Test Plan
<!-- Describe how you tested this functionality. Include steps to reproduce, relevant test cases, and any other pertinent information. -->
### PR Quality Checklist
- [ ] My pull request is atomic and focuses on a single change.
- [ ] I have thoroughly tested my changes with multiple different prompts.
- [ ] I have considered potential risks and mitigations for my changes.
- [ ] I have documented my changes clearly and comprehensively.
- [ ] I have not snuck in any "extra" small tweaks changes. <!-- Submit these as separate Pull Requests, they are the easiest to merge! -->
- [ ] I have run the following commands against my code to ensure it passes our linters:
```shell
black .
isort .
mypy
autoflake --remove-all-unused-imports --recursive --ignore-init-module-imports --ignore-pass-after-docstring autogpt tests --in-place
```
<!-- If you haven't added tests, please explain why. If you have, check the appropriate box. If you've ensured your PR is atomic and well-documented, check the corresponding boxes. -->
<!-- By submitting this, I agree that my pull request should be closed if I do not fill this out or follow the guidelines. -->

27
.github/labeler.yml vendored
View File

@@ -1,27 +0,0 @@
AutoGPT Agent:
- changed-files:
- any-glob-to-any-file: classic/original_autogpt/**
Forge:
- changed-files:
- any-glob-to-any-file: classic/forge/**
Benchmark:
- changed-files:
- any-glob-to-any-file: classic/benchmark/**
Frontend:
- changed-files:
- any-glob-to-any-file: classic/frontend/**
documentation:
- changed-files:
- any-glob-to-any-file: docs/**
Builder:
- changed-files:
- any-glob-to-any-file: autogpt_platform/autogpt_builder/**
Server:
- changed-files:
- any-glob-to-any-file: autogpt_platform/autogpt_server/**

73
.github/workflows/benchmarks.yml vendored Normal file
View File

@@ -0,0 +1,73 @@
name: Benchmarks
on:
schedule:
- cron: '0 8 * * *'
workflow_dispatch:
jobs:
Benchmark:
name: ${{ matrix.config.task-name }}
runs-on: ubuntu-latest
timeout-minutes: 30
strategy:
fail-fast: false
matrix:
config:
- python-version: "3.10"
task: "tests/challenges"
task-name: "Mandatory Tasks"
- python-version: "3.10"
task: "--beat-challenges -ra tests/challenges"
task-name: "Challenging Tasks"
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
ref: master
- name: Set up Python ${{ matrix.config.python-version }}
uses: actions/setup-python@v4
with:
python-version: ${{ matrix.config.python-version }}
- id: get_date
name: Get date
run: echo "date=$(date +'%Y-%m-%d')" >> $GITHUB_OUTPUT
- name: Set up Python dependency cache
uses: actions/cache@v3
with:
path: ~/.cache/pip
key: ${{ runner.os }}-pip-${{ hashFiles('**/requirements.txt') }}-${{ steps.get_date.outputs.date }}
- name: Install Python dependencies
run: |
python -m pip install --upgrade pip
pip install -r requirements.txt
- name: Run pytest with coverage
run: |
rm -rf tests/Auto-GPT-test-cassettes
pytest -n auto --record-mode=all ${{ matrix.config.task }}
env:
CI: true
PROXY: ${{ secrets.PROXY }}
AGENT_MODE: ${{ secrets.AGENT_MODE }}
AGENT_TYPE: ${{ secrets.AGENT_TYPE }}
PLAIN_OUTPUT: True
- name: Upload logs as artifact
if: always()
uses: actions/upload-artifact@v3
with:
name: test-logs-${{ matrix.config.task-name }}
path: logs/
- name: Upload cassettes as artifact
if: always()
uses: actions/upload-artifact@v3
with:
name: cassettes-${{ matrix.config.task-name }}
path: tests/Auto-GPT-test-cassettes/

261
.github/workflows/ci.yml vendored Normal file
View File

@@ -0,0 +1,261 @@
name: Python CI
on:
push:
branches: [ master, ci-test* ]
paths-ignore:
- 'tests/Auto-GPT-test-cassettes'
- 'tests/challenges/current_score.json'
pull_request:
branches: [ stable, master, release-* ]
pull_request_target:
branches: [ master, release-*, ci-test* ]
concurrency:
group: ${{ format('ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ startsWith(github.event_name, 'pull_request') }}
jobs:
lint:
# eliminate duplicate runs
if: github.event_name == 'push' || (github.event.pull_request.head.repo.fork == (github.event_name == 'pull_request_target'))
runs-on: ubuntu-latest
env:
min-python-version: "3.10"
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
fetch-depth: 0
ref: ${{ github.event.pull_request.head.ref }}
repository: ${{ github.event.pull_request.head.repo.full_name }}
- name: Set up Python ${{ env.min-python-version }}
uses: actions/setup-python@v4
with:
python-version: ${{ env.min-python-version }}
- id: get_date
name: Get date
run: echo "date=$(date +'%Y-%m-%d')" >> $GITHUB_OUTPUT
- name: Set up Python dependency cache
uses: actions/cache@v3
with:
path: ~/.cache/pip
key: ${{ runner.os }}-pip-${{ hashFiles('**/requirements.txt') }}-${{ steps.get_date.outputs.date }}
- name: Install dependencies
run: |
python -m pip install --upgrade pip
pip install -r requirements.txt
- name: Lint with flake8
run: flake8
- name: Check black formatting
run: black . --check
if: success() || failure()
- name: Check isort formatting
run: isort . --check
if: success() || failure()
- name: Check mypy formatting
run: mypy
if: success() || failure()
- name: Check for unused imports and pass statements
run: |
cmd="autoflake --remove-all-unused-imports --recursive --ignore-init-module-imports --ignore-pass-after-docstring autogpt tests"
$cmd --check || (echo "You have unused imports or pass statements, please run '${cmd} --in-place'" && exit 1)
test:
# eliminate duplicate runs
if: github.event_name == 'push' || (github.event.pull_request.head.repo.fork == (github.event_name == 'pull_request_target'))
permissions:
# Gives the action the necessary permissions for publishing new
# comments in pull requests.
pull-requests: write
# Gives the action the necessary permissions for pushing data to the
# python-coverage-comment-action branch, and for editing existing
# comments (to avoid publishing multiple comments in the same PR)
contents: write
runs-on: ubuntu-latest
timeout-minutes: 30
strategy:
matrix:
python-version: ["3.10"]
steps:
- name: Checkout repository
uses: actions/checkout@v3
with:
fetch-depth: 0
ref: ${{ github.event.pull_request.head.ref }}
repository: ${{ github.event.pull_request.head.repo.full_name }}
submodules: true
- name: Configure git user Auto-GPT-Bot
run: |
git config --global user.name "Auto-GPT-Bot"
git config --global user.email "github-bot@agpt.co"
- name: Checkout cassettes
if: ${{ startsWith(github.event_name, 'pull_request') }}
run: |
cassette_branch="${{ github.event.pull_request.user.login }}-${{ github.event.pull_request.head.ref }}"
cassette_base_branch="${{ github.event.pull_request.base.ref }}"
cd tests/Auto-GPT-test-cassettes
if ! git ls-remote --exit-code --heads origin $cassette_base_branch ; then
cassette_base_branch="master"
fi
if git ls-remote --exit-code --heads origin $cassette_branch ; then
git fetch origin $cassette_branch
git fetch origin $cassette_base_branch
git checkout $cassette_branch
# Pick non-conflicting cassette updates from the base branch
git merge --no-commit --strategy-option=ours origin/$cassette_base_branch
echo "Using cassettes from mirror branch '$cassette_branch'," \
"synced to upstream branch '$cassette_base_branch'."
else
git checkout -b $cassette_branch
echo "Branch '$cassette_branch' does not exist in cassette submodule." \
"Using cassettes from '$cassette_base_branch'."
fi
- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v4
with:
python-version: ${{ matrix.python-version }}
- id: get_date
name: Get date
run: echo "date=$(date +'%Y-%m-%d')" >> $GITHUB_OUTPUT
- name: Set up Python dependency cache
uses: actions/cache@v3
with:
path: ~/.cache/pip
key: ${{ runner.os }}-pip-${{ hashFiles('**/requirements.txt') }}-${{ steps.get_date.outputs.date }}
- name: Install Python dependencies
run: |
python -m pip install --upgrade pip
pip install -r requirements.txt
- name: Run pytest with coverage
run: |
pytest -vv --cov=autogpt --cov-branch --cov-report term-missing --cov-report xml \
--numprocesses=logical --durations=10 \
tests/unit tests/integration tests/challenges
python tests/challenges/utils/build_current_score.py
env:
CI: true
PROXY: ${{ github.event_name == 'pull_request_target' && secrets.PROXY || '' }}
AGENT_MODE: ${{ github.event_name == 'pull_request_target' && secrets.AGENT_MODE || '' }}
AGENT_TYPE: ${{ github.event_name == 'pull_request_target' && secrets.AGENT_TYPE || '' }}
OPENAI_API_KEY: ${{ github.event_name != 'pull_request_target' && secrets.OPENAI_API_KEY || '' }}
PLAIN_OUTPUT: True
- name: Upload coverage reports to Codecov
uses: codecov/codecov-action@v3
- id: setup_git_auth
name: Set up git token authentication
# Cassettes may be pushed even when tests fail
if: success() || failure()
run: |
config_key="http.${{ github.server_url }}/.extraheader"
base64_pat=$(echo -n "pat:${{ secrets.PAT_REVIEW }}" | base64 -w0)
git config "$config_key" \
"Authorization: Basic $base64_pat"
cd tests/Auto-GPT-test-cassettes
git config "$config_key" \
"Authorization: Basic $base64_pat"
echo "config_key=$config_key" >> $GITHUB_OUTPUT
- name: Push updated challenge scores
if: github.event_name == 'push'
run: |
score_file="tests/challenges/current_score.json"
if ! git diff --quiet $score_file; then
git add $score_file
git commit -m "Update challenge scores"
git push origin HEAD:${{ github.ref_name }}
else
echo "The challenge scores didn't change."
fi
- id: push_cassettes
name: Push updated cassettes
# For pull requests, push updated cassettes even when tests fail
if: github.event_name == 'push' || success() || failure()
run: |
if [ "${{ startsWith(github.event_name, 'pull_request') }}" = "true" ]; then
is_pull_request=true
cassette_branch="${{ github.event.pull_request.user.login }}-${{ github.event.pull_request.head.ref }}"
else
cassette_branch="${{ github.ref_name }}"
fi
cd tests/Auto-GPT-test-cassettes
# Commit & push changes to cassettes if any
if ! git diff --quiet; then
git add .
git commit -m "Auto-update cassettes"
git push origin HEAD:$cassette_branch
if [ ! $is_pull_request ]; then
cd ../..
git add tests/Auto-GPT-test-cassettes
git commit -m "Update cassette submodule"
git push origin HEAD:$cassette_branch
fi
echo "updated=true" >> $GITHUB_OUTPUT
else
echo "updated=false" >> $GITHUB_OUTPUT
echo "No cassette changes to commit"
fi
- name: Post Set up git token auth
if: steps.setup_git_auth.outcome == 'success'
run: |
git config --unset-all '${{ steps.setup_git_auth.outputs.config_key }}'
git submodule foreach git config --unset-all '${{ steps.setup_git_auth.outputs.config_key }}'
- name: Apply "behaviour change" label and comment on PR
if: ${{ startsWith(github.event_name, 'pull_request') }}
run: |
PR_NUMBER=${{ github.event.pull_request.number }}
TOKEN=${{ secrets.PAT_REVIEW }}
REPO=${{ github.repository }}
if [[ "${{ steps.push_cassettes.outputs.updated }}" == "true" ]]; then
echo "Adding label and comment..."
curl -X POST \
-H "Authorization: Bearer $TOKEN" \
-H "Accept: application/vnd.github.v3+json" \
https://api.github.com/repos/$REPO/issues/$PR_NUMBER/labels \
-d '{"labels":["behaviour change"]}'
echo $TOKEN | gh auth login --with-token
gh api repos/$REPO/issues/$PR_NUMBER/comments -X POST -F body="You changed AutoGPT's behaviour. The cassettes have been updated and will be merged to the submodule when this Pull Request gets merged."
fi
- name: Upload logs to artifact
if: always()
uses: actions/upload-artifact@v3
with:
name: test-logs
path: logs/

View File

@@ -1,138 +0,0 @@
name: Classic - AutoGPT CI
on:
push:
branches: [ master, development, ci-test* ]
paths:
- '.github/workflows/classic-autogpt-ci.yml'
- 'classic/original_autogpt/**'
pull_request:
branches: [ master, development, release-* ]
paths:
- '.github/workflows/classic-autogpt-ci.yml'
- 'classic/original_autogpt/**'
concurrency:
group: ${{ format('classic-autogpt-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ startsWith(github.event_name, 'pull_request') }}
defaults:
run:
shell: bash
working-directory: classic/original_autogpt
jobs:
test:
permissions:
contents: read
timeout-minutes: 30
strategy:
fail-fast: false
matrix:
python-version: ["3.10"]
platform-os: [ubuntu, macos, macos-arm64, windows]
runs-on: ${{ matrix.platform-os != 'macos-arm64' && format('{0}-latest', matrix.platform-os) || 'macos-14' }}
steps:
# Quite slow on macOS (2~4 minutes to set up Docker)
# - name: Set up Docker (macOS)
# if: runner.os == 'macOS'
# uses: crazy-max/ghaction-setup-docker@v3
- name: Start MinIO service (Linux)
if: runner.os == 'Linux'
working-directory: '.'
run: |
docker pull minio/minio:edge-cicd
docker run -d -p 9000:9000 minio/minio:edge-cicd
- name: Start MinIO service (macOS)
if: runner.os == 'macOS'
working-directory: ${{ runner.temp }}
run: |
brew install minio/stable/minio
mkdir data
minio server ./data &
# No MinIO on Windows:
# - Windows doesn't support running Linux Docker containers
# - It doesn't seem possible to start background processes on Windows. They are
# killed after the step returns.
# See: https://github.com/actions/runner/issues/598#issuecomment-2011890429
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
submodules: true
- name: Configure git user Auto-GPT-Bot
run: |
git config --global user.name "Auto-GPT-Bot"
git config --global user.email "github-bot@agpt.co"
- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
- id: get_date
name: Get date
run: echo "date=$(date +'%Y-%m-%d')" >> $GITHUB_OUTPUT
- name: Set up Python dependency cache
# On Windows, unpacking cached dependencies takes longer than just installing them
if: runner.os != 'Windows'
uses: actions/cache@v4
with:
path: ${{ runner.os == 'macOS' && '~/Library/Caches/pypoetry' || '~/.cache/pypoetry' }}
key: poetry-${{ runner.os }}-${{ hashFiles('classic/original_autogpt/poetry.lock') }}
- name: Install Poetry (Unix)
if: runner.os != 'Windows'
run: |
curl -sSL https://install.python-poetry.org | python3 -
if [ "${{ runner.os }}" = "macOS" ]; then
PATH="$HOME/.local/bin:$PATH"
echo "$HOME/.local/bin" >> $GITHUB_PATH
fi
- name: Install Poetry (Windows)
if: runner.os == 'Windows'
shell: pwsh
run: |
(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -
$env:PATH += ";$env:APPDATA\Python\Scripts"
echo "$env:APPDATA\Python\Scripts" >> $env:GITHUB_PATH
- name: Install Python dependencies
run: poetry install
- name: Run pytest with coverage
run: |
poetry run pytest -vv \
--cov=autogpt --cov-branch --cov-report term-missing --cov-report xml \
--numprocesses=logical --durations=10 \
tests/unit tests/integration
env:
CI: true
PLAIN_OUTPUT: True
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
S3_ENDPOINT_URL: ${{ runner.os != 'Windows' && 'http://127.0.0.1:9000' || '' }}
AWS_ACCESS_KEY_ID: minioadmin
AWS_SECRET_ACCESS_KEY: minioadmin
- name: Upload coverage reports to Codecov
uses: codecov/codecov-action@v4
with:
token: ${{ secrets.CODECOV_TOKEN }}
flags: autogpt-agent,${{ runner.os }}
- name: Upload logs to artifact
if: always()
uses: actions/upload-artifact@v4
with:
name: test-logs
path: classic/original_autogpt/logs/

View File

@@ -1,166 +0,0 @@
name: Classic - AutoGPT Docker CI
on:
push:
branches: [ master, development ]
paths:
- '.github/workflows/classic-autogpt-docker-ci.yml'
- 'classic/original_autogpt/**'
- 'classic/forge/**'
pull_request:
branches: [ master, development, release-* ]
paths:
- '.github/workflows/classic-autogpt-docker-ci.yml'
- 'classic/original_autogpt/**'
- 'classic/forge/**'
concurrency:
group: ${{ format('classic-autogpt-docker-ci-{0}', github.head_ref && format('pr-{0}', github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ github.event_name == 'pull_request' }}
defaults:
run:
working-directory: classic/original_autogpt
env:
IMAGE_NAME: auto-gpt
DEPLOY_IMAGE_NAME: ${{ secrets.DOCKER_USER && format('{0}/', secrets.DOCKER_USER) || '' }}auto-gpt
DEV_IMAGE_TAG: latest-dev
jobs:
build:
runs-on: ubuntu-latest
strategy:
matrix:
build-type: [release, dev]
steps:
- name: Checkout repository
uses: actions/checkout@v4
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
- if: runner.debug
run: |
ls -al
du -hs *
- id: build
name: Build image
uses: docker/build-push-action@v5
with:
context: classic/
file: classic/Dockerfile.autogpt
build-args: BUILD_TYPE=${{ matrix.build-type }}
tags: ${{ env.IMAGE_NAME }}
labels: GIT_REVISION=${{ github.sha }}
load: true # save to docker images
# cache layers in GitHub Actions cache to speed up builds
cache-from: type=gha,scope=autogpt-docker-${{ matrix.build-type }}
cache-to: type=gha,scope=autogpt-docker-${{ matrix.build-type }},mode=max
- name: Generate build report
env:
event_name: ${{ github.event_name }}
event_ref: ${{ github.event.ref }}
event_ref_type: ${{ github.event.ref}}
build_type: ${{ matrix.build-type }}
prod_branch: master
dev_branch: development
repository: ${{ github.repository }}
base_branch: ${{ github.ref_name != 'master' && github.ref_name != 'development' && 'development' || 'master' }}
current_ref: ${{ github.ref_name }}
commit_hash: ${{ github.event.after }}
source_url: ${{ format('{0}/tree/{1}', github.event.repository.url, github.event.release && github.event.release.tag_name || github.sha) }}
push_forced_label: ${{ github.event.forced && '☢️ forced' || '' }}
new_commits_json: ${{ toJSON(github.event.commits) }}
compare_url_template: ${{ format('/{0}/compare/{{base}}...{{head}}', github.repository) }}
github_context_json: ${{ toJSON(github) }}
job_env_json: ${{ toJSON(env) }}
vars_json: ${{ toJSON(vars) }}
run: .github/workflows/scripts/docker-ci-summary.sh >> $GITHUB_STEP_SUMMARY
continue-on-error: true
test:
runs-on: ubuntu-latest
timeout-minutes: 10
services:
minio:
image: minio/minio:edge-cicd
options: >
--name=minio
--health-interval=10s --health-timeout=5s --health-retries=3
--health-cmd="curl -f http://localhost:9000/minio/health/live"
steps:
- name: Check out repository
uses: actions/checkout@v4
with:
submodules: true
- if: github.event_name == 'push'
name: Log in to Docker hub
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKER_USER }}
password: ${{ secrets.DOCKER_PASSWORD }}
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
- id: build
name: Build image
uses: docker/build-push-action@v5
with:
context: classic/
file: classic/Dockerfile.autogpt
build-args: BUILD_TYPE=dev # include pytest
tags: >
${{ env.IMAGE_NAME }},
${{ env.DEPLOY_IMAGE_NAME }}:${{ env.DEV_IMAGE_TAG }}
labels: GIT_REVISION=${{ github.sha }}
load: true # save to docker images
# cache layers in GitHub Actions cache to speed up builds
cache-from: type=gha,scope=autogpt-docker-dev
cache-to: type=gha,scope=autogpt-docker-dev,mode=max
- id: test
name: Run tests
env:
CI: true
PLAIN_OUTPUT: True
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
S3_ENDPOINT_URL: http://minio:9000
AWS_ACCESS_KEY_ID: minioadmin
AWS_SECRET_ACCESS_KEY: minioadmin
run: |
set +e
docker run --env CI --env OPENAI_API_KEY \
--network container:minio \
--env S3_ENDPOINT_URL --env AWS_ACCESS_KEY_ID --env AWS_SECRET_ACCESS_KEY \
--entrypoint poetry ${{ env.IMAGE_NAME }} run \
pytest -v --cov=autogpt --cov-branch --cov-report term-missing \
--numprocesses=4 --durations=10 \
tests/unit tests/integration 2>&1 | tee test_output.txt
test_failure=${PIPESTATUS[0]}
cat << $EOF >> $GITHUB_STEP_SUMMARY
# Tests $([ $test_failure = 0 ] && echo '✅' || echo '❌')
\`\`\`
$(cat test_output.txt)
\`\`\`
$EOF
exit $test_failure
- if: github.event_name == 'push' && github.ref_name == 'master'
name: Push image to Docker Hub
run: docker push ${{ env.DEPLOY_IMAGE_NAME }}:${{ env.DEV_IMAGE_TAG }}

View File

@@ -1,76 +0,0 @@
name: Classic - Agent smoke tests
on:
workflow_dispatch:
schedule:
- cron: '0 8 * * *'
push:
branches: [ master, development, ci-test* ]
paths:
- '.github/workflows/classic-autogpts-ci.yml'
- 'classic/original_autogpt/**'
- 'classic/forge/**'
- 'classic/benchmark/**'
- 'classic/run'
- 'classic/cli.py'
- 'classic/setup.py'
- '!**/*.md'
pull_request:
branches: [ master, development, release-* ]
paths:
- '.github/workflows/classic-autogpts-ci.yml'
- 'classic/original_autogpt/**'
- 'classic/forge/**'
- 'classic/benchmark/**'
- 'classic/run'
- 'classic/cli.py'
- 'classic/setup.py'
- '!**/*.md'
defaults:
run:
shell: bash
working-directory: classic
jobs:
serve-agent-protocol:
runs-on: ubuntu-latest
strategy:
matrix:
agent-name: [ original_autogpt ]
fail-fast: false
timeout-minutes: 20
env:
min-python-version: '3.10'
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
submodules: true
- name: Set up Python ${{ env.min-python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ env.min-python-version }}
- name: Install Poetry
working-directory: ./classic/${{ matrix.agent-name }}/
run: |
curl -sSL https://install.python-poetry.org | python -
- name: Run regression tests
run: |
./run agent start ${{ matrix.agent-name }}
cd ${{ matrix.agent-name }}
poetry run agbenchmark --mock --test=BasicRetrieval --test=Battleship --test=WebArenaTask_0
poetry run agbenchmark --test=WriteFile
env:
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
AGENT_NAME: ${{ matrix.agent-name }}
REQUESTS_CA_BUNDLE: /etc/ssl/certs/ca-certificates.crt
HELICONE_CACHE_ENABLED: false
HELICONE_PROPERTY_AGENT: ${{ matrix.agent-name }}
REPORTS_FOLDER: ${{ format('../../reports/{0}', matrix.agent-name) }}
TELEMETRY_ENVIRONMENT: autogpt-ci
TELEMETRY_OPT_IN: ${{ github.ref_name == 'master' }}

View File

@@ -1,169 +0,0 @@
name: Classic - AGBenchmark CI
on:
push:
branches: [ master, development, ci-test* ]
paths:
- 'classic/benchmark/**'
- '!classic/benchmark/reports/**'
- .github/workflows/classic-benchmark-ci.yml
pull_request:
branches: [ master, development, release-* ]
paths:
- 'classic/benchmark/**'
- '!classic/benchmark/reports/**'
- .github/workflows/classic-benchmark-ci.yml
concurrency:
group: ${{ format('benchmark-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ startsWith(github.event_name, 'pull_request') }}
defaults:
run:
shell: bash
env:
min-python-version: '3.10'
jobs:
test:
permissions:
contents: read
timeout-minutes: 30
strategy:
fail-fast: false
matrix:
python-version: ["3.10"]
platform-os: [ubuntu, macos, macos-arm64, windows]
runs-on: ${{ matrix.platform-os != 'macos-arm64' && format('{0}-latest', matrix.platform-os) || 'macos-14' }}
defaults:
run:
shell: bash
working-directory: classic/benchmark
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
submodules: true
- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
- name: Set up Python dependency cache
# On Windows, unpacking cached dependencies takes longer than just installing them
if: runner.os != 'Windows'
uses: actions/cache@v4
with:
path: ${{ runner.os == 'macOS' && '~/Library/Caches/pypoetry' || '~/.cache/pypoetry' }}
key: poetry-${{ runner.os }}-${{ hashFiles('classic/benchmark/poetry.lock') }}
- name: Install Poetry (Unix)
if: runner.os != 'Windows'
run: |
curl -sSL https://install.python-poetry.org | python3 -
if [ "${{ runner.os }}" = "macOS" ]; then
PATH="$HOME/.local/bin:$PATH"
echo "$HOME/.local/bin" >> $GITHUB_PATH
fi
- name: Install Poetry (Windows)
if: runner.os == 'Windows'
shell: pwsh
run: |
(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -
$env:PATH += ";$env:APPDATA\Python\Scripts"
echo "$env:APPDATA\Python\Scripts" >> $env:GITHUB_PATH
- name: Install Python dependencies
run: poetry install
- name: Run pytest with coverage
run: |
poetry run pytest -vv \
--cov=agbenchmark --cov-branch --cov-report term-missing --cov-report xml \
--durations=10 \
tests
env:
CI: true
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
- name: Upload coverage reports to Codecov
uses: codecov/codecov-action@v4
with:
token: ${{ secrets.CODECOV_TOKEN }}
flags: agbenchmark,${{ runner.os }}
self-test-with-agent:
runs-on: ubuntu-latest
strategy:
matrix:
agent-name: [ forge ]
fail-fast: false
timeout-minutes: 20
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
submodules: true
- name: Set up Python ${{ env.min-python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ env.min-python-version }}
- name: Install Poetry
run: |
curl -sSL https://install.python-poetry.org | python -
- name: Run regression tests
working-directory: classic
run: |
./run agent start ${{ matrix.agent-name }}
cd ${{ matrix.agent-name }}
set +e # Ignore non-zero exit codes and continue execution
echo "Running the following command: poetry run agbenchmark --maintain --mock"
poetry run agbenchmark --maintain --mock
EXIT_CODE=$?
set -e # Stop ignoring non-zero exit codes
# Check if the exit code was 5, and if so, exit with 0 instead
if [ $EXIT_CODE -eq 5 ]; then
echo "regression_tests.json is empty."
fi
echo "Running the following command: poetry run agbenchmark --mock"
poetry run agbenchmark --mock
echo "Running the following command: poetry run agbenchmark --mock --category=data"
poetry run agbenchmark --mock --category=data
echo "Running the following command: poetry run agbenchmark --mock --category=coding"
poetry run agbenchmark --mock --category=coding
echo "Running the following command: poetry run agbenchmark --test=WriteFile"
poetry run agbenchmark --test=WriteFile
cd ../benchmark
poetry install
echo "Adding the BUILD_SKILL_TREE environment variable. This will attempt to add new elements in the skill tree. If new elements are added, the CI fails because they should have been pushed"
export BUILD_SKILL_TREE=true
poetry run agbenchmark --mock
CHANGED=$(git diff --name-only | grep -E '(agclassic/benchmark/challenges)|(../classic/frontend/assets)') || echo "No diffs"
if [ ! -z "$CHANGED" ]; then
echo "There are unstaged changes please run agbenchmark and commit those changes since they are needed."
echo "$CHANGED"
exit 1
else
echo "No unstaged changes."
fi
env:
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
TELEMETRY_ENVIRONMENT: autogpt-benchmark-ci
TELEMETRY_OPT_IN: ${{ github.ref_name == 'master' }}

View File

@@ -1,55 +0,0 @@
name: Classic - Publish to PyPI
on:
workflow_dispatch:
jobs:
deploy:
runs-on: ubuntu-latest
permissions:
contents: write
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
submodules: true
fetch-depth: 0
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: 3.8
- name: Install Poetry
working-directory: ./classic/benchmark/
run: |
curl -sSL https://install.python-poetry.org | python3 -
echo "$HOME/.poetry/bin" >> $GITHUB_PATH
- name: Build project for distribution
working-directory: ./classic/benchmark/
run: poetry build
- name: Install dependencies
working-directory: ./classic/benchmark/
run: poetry install
- name: Check Version
working-directory: ./classic/benchmark/
id: check-version
run: |
echo version=$(poetry version --short) >> $GITHUB_OUTPUT
- name: Create Release
uses: ncipollo/release-action@v1
with:
artifacts: "classic/benchmark/dist/*"
token: ${{ secrets.GITHUB_TOKEN }}
draft: false
generateReleaseNotes: false
tag: agbenchmark-v${{ steps.check-version.outputs.version }}
commit: master
- name: Build and publish
working-directory: ./classic/benchmark/
run: poetry publish -u __token__ -p ${{ secrets.PYPI_API_TOKEN }}

View File

@@ -1,236 +0,0 @@
name: Classic - Forge CI
on:
push:
branches: [ master, development, ci-test* ]
paths:
- '.github/workflows/classic-forge-ci.yml'
- 'classic/forge/**'
- '!classic/forge/tests/vcr_cassettes'
pull_request:
branches: [ master, development, release-* ]
paths:
- '.github/workflows/classic-forge-ci.yml'
- 'classic/forge/**'
- '!classic/forge/tests/vcr_cassettes'
concurrency:
group: ${{ format('forge-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ startsWith(github.event_name, 'pull_request') }}
defaults:
run:
shell: bash
working-directory: classic/forge
jobs:
test:
permissions:
contents: read
timeout-minutes: 30
strategy:
fail-fast: false
matrix:
python-version: ["3.10"]
platform-os: [ubuntu, macos, macos-arm64, windows]
runs-on: ${{ matrix.platform-os != 'macos-arm64' && format('{0}-latest', matrix.platform-os) || 'macos-14' }}
steps:
# Quite slow on macOS (2~4 minutes to set up Docker)
# - name: Set up Docker (macOS)
# if: runner.os == 'macOS'
# uses: crazy-max/ghaction-setup-docker@v3
- name: Start MinIO service (Linux)
if: runner.os == 'Linux'
working-directory: '.'
run: |
docker pull minio/minio:edge-cicd
docker run -d -p 9000:9000 minio/minio:edge-cicd
- name: Start MinIO service (macOS)
if: runner.os == 'macOS'
working-directory: ${{ runner.temp }}
run: |
brew install minio/stable/minio
mkdir data
minio server ./data &
# No MinIO on Windows:
# - Windows doesn't support running Linux Docker containers
# - It doesn't seem possible to start background processes on Windows. They are
# killed after the step returns.
# See: https://github.com/actions/runner/issues/598#issuecomment-2011890429
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
submodules: true
- name: Checkout cassettes
if: ${{ startsWith(github.event_name, 'pull_request') }}
env:
PR_BASE: ${{ github.event.pull_request.base.ref }}
PR_BRANCH: ${{ github.event.pull_request.head.ref }}
PR_AUTHOR: ${{ github.event.pull_request.user.login }}
run: |
cassette_branch="${PR_AUTHOR}-${PR_BRANCH}"
cassette_base_branch="${PR_BASE}"
cd tests/vcr_cassettes
if ! git ls-remote --exit-code --heads origin $cassette_base_branch ; then
cassette_base_branch="master"
fi
if git ls-remote --exit-code --heads origin $cassette_branch ; then
git fetch origin $cassette_branch
git fetch origin $cassette_base_branch
git checkout $cassette_branch
# Pick non-conflicting cassette updates from the base branch
git merge --no-commit --strategy-option=ours origin/$cassette_base_branch
echo "Using cassettes from mirror branch '$cassette_branch'," \
"synced to upstream branch '$cassette_base_branch'."
else
git checkout -b $cassette_branch
echo "Branch '$cassette_branch' does not exist in cassette submodule." \
"Using cassettes from '$cassette_base_branch'."
fi
- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
- name: Set up Python dependency cache
# On Windows, unpacking cached dependencies takes longer than just installing them
if: runner.os != 'Windows'
uses: actions/cache@v4
with:
path: ${{ runner.os == 'macOS' && '~/Library/Caches/pypoetry' || '~/.cache/pypoetry' }}
key: poetry-${{ runner.os }}-${{ hashFiles('classic/forge/poetry.lock') }}
- name: Install Poetry (Unix)
if: runner.os != 'Windows'
run: |
curl -sSL https://install.python-poetry.org | python3 -
if [ "${{ runner.os }}" = "macOS" ]; then
PATH="$HOME/.local/bin:$PATH"
echo "$HOME/.local/bin" >> $GITHUB_PATH
fi
- name: Install Poetry (Windows)
if: runner.os == 'Windows'
shell: pwsh
run: |
(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -
$env:PATH += ";$env:APPDATA\Python\Scripts"
echo "$env:APPDATA\Python\Scripts" >> $env:GITHUB_PATH
- name: Install Python dependencies
run: poetry install
- name: Run pytest with coverage
run: |
poetry run pytest -vv \
--cov=forge --cov-branch --cov-report term-missing --cov-report xml \
--durations=10 \
forge
env:
CI: true
PLAIN_OUTPUT: True
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
S3_ENDPOINT_URL: ${{ runner.os != 'Windows' && 'http://127.0.0.1:9000' || '' }}
AWS_ACCESS_KEY_ID: minioadmin
AWS_SECRET_ACCESS_KEY: minioadmin
- name: Upload coverage reports to Codecov
uses: codecov/codecov-action@v4
with:
token: ${{ secrets.CODECOV_TOKEN }}
flags: forge,${{ runner.os }}
- id: setup_git_auth
name: Set up git token authentication
# Cassettes may be pushed even when tests fail
if: success() || failure()
run: |
config_key="http.${{ github.server_url }}/.extraheader"
if [ "${{ runner.os }}" = 'macOS' ]; then
base64_pat=$(echo -n "pat:${{ secrets.PAT_REVIEW }}" | base64)
else
base64_pat=$(echo -n "pat:${{ secrets.PAT_REVIEW }}" | base64 -w0)
fi
git config "$config_key" \
"Authorization: Basic $base64_pat"
cd tests/vcr_cassettes
git config "$config_key" \
"Authorization: Basic $base64_pat"
echo "config_key=$config_key" >> $GITHUB_OUTPUT
- id: push_cassettes
name: Push updated cassettes
# For pull requests, push updated cassettes even when tests fail
if: github.event_name == 'push' || (! github.event.pull_request.head.repo.fork && (success() || failure()))
env:
PR_BRANCH: ${{ github.event.pull_request.head.ref }}
PR_AUTHOR: ${{ github.event.pull_request.user.login }}
run: |
if [ "${{ startsWith(github.event_name, 'pull_request') }}" = "true" ]; then
is_pull_request=true
cassette_branch="${PR_AUTHOR}-${PR_BRANCH}"
else
cassette_branch="${{ github.ref_name }}"
fi
cd tests/vcr_cassettes
# Commit & push changes to cassettes if any
if ! git diff --quiet; then
git add .
git commit -m "Auto-update cassettes"
git push origin HEAD:$cassette_branch
if [ ! $is_pull_request ]; then
cd ../..
git add tests/vcr_cassettes
git commit -m "Update cassette submodule"
git push origin HEAD:$cassette_branch
fi
echo "updated=true" >> $GITHUB_OUTPUT
else
echo "updated=false" >> $GITHUB_OUTPUT
echo "No cassette changes to commit"
fi
- name: Post Set up git token auth
if: steps.setup_git_auth.outcome == 'success'
run: |
git config --unset-all '${{ steps.setup_git_auth.outputs.config_key }}'
git submodule foreach git config --unset-all '${{ steps.setup_git_auth.outputs.config_key }}'
- name: Apply "behaviour change" label and comment on PR
if: ${{ startsWith(github.event_name, 'pull_request') }}
run: |
PR_NUMBER="${{ github.event.pull_request.number }}"
TOKEN="${{ secrets.PAT_REVIEW }}"
REPO="${{ github.repository }}"
if [[ "${{ steps.push_cassettes.outputs.updated }}" == "true" ]]; then
echo "Adding label and comment..."
echo $TOKEN | gh auth login --with-token
gh issue edit $PR_NUMBER --add-label "behaviour change"
gh issue comment $PR_NUMBER --body "You changed AutoGPT's behaviour on ${{ runner.os }}. The cassettes have been updated and will be merged to the submodule when this Pull Request gets merged."
fi
- name: Upload logs to artifact
if: always()
uses: actions/upload-artifact@v4
with:
name: test-logs
path: classic/forge/logs/

View File

@@ -1,60 +0,0 @@
name: Classic - Frontend CI/CD
on:
push:
branches:
- master
- development
- 'ci-test*' # This will match any branch that starts with "ci-test"
paths:
- 'classic/frontend/**'
- '.github/workflows/frontend-ci.yml'
pull_request:
paths:
- 'classic/frontend/**'
- '.github/workflows/frontend-ci.yml'
jobs:
build:
permissions:
contents: write
pull-requests: write
runs-on: ubuntu-latest
env:
BUILD_BRANCH: ${{ format('frontend-build/{0}', github.ref_name) }}
steps:
- name: Checkout Repo
uses: actions/checkout@v4
- name: Setup Flutter
uses: subosito/flutter-action@v2
with:
flutter-version: '3.13.2'
- name: Build Flutter to Web
run: |
cd classic/frontend
flutter build web --base-href /app/
# - name: Commit and Push to ${{ env.BUILD_BRANCH }}
# if: github.event_name == 'push'
# run: |
# git config --local user.email "action@github.com"
# git config --local user.name "GitHub Action"
# git add classic/frontend/build/web
# git checkout -B ${{ env.BUILD_BRANCH }}
# git commit -m "Update frontend build to ${GITHUB_SHA:0:7}" -a
# git push -f origin ${{ env.BUILD_BRANCH }}
- name: Create PR ${{ env.BUILD_BRANCH }} -> ${{ github.ref_name }}
if: github.event_name == 'push'
uses: peter-evans/create-pull-request@v6
with:
add-paths: classic/frontend/build/web
base: ${{ github.ref_name }}
branch: ${{ env.BUILD_BRANCH }}
delete-branch: true
title: "Update frontend build in `${{ github.ref_name }}`"
body: "This PR updates the frontend build based on commit ${{ github.sha }}."
commit-message: "Update frontend build based on commit ${{ github.sha }}"

View File

@@ -1,151 +0,0 @@
name: Classic - Python checks
on:
push:
branches: [ master, development, ci-test* ]
paths:
- '.github/workflows/lint-ci.yml'
- 'classic/original_autogpt/**'
- 'classic/forge/**'
- 'classic/benchmark/**'
- '**.py'
- '!classic/forge/tests/vcr_cassettes'
pull_request:
branches: [ master, development, release-* ]
paths:
- '.github/workflows/lint-ci.yml'
- 'classic/original_autogpt/**'
- 'classic/forge/**'
- 'classic/benchmark/**'
- '**.py'
- '!classic/forge/tests/vcr_cassettes'
concurrency:
group: ${{ format('lint-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ startsWith(github.event_name, 'pull_request') }}
defaults:
run:
shell: bash
jobs:
get-changed-parts:
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v4
- id: changes-in
name: Determine affected subprojects
uses: dorny/paths-filter@v3
with:
filters: |
original_autogpt:
- classic/original_autogpt/autogpt/**
- classic/original_autogpt/tests/**
- classic/original_autogpt/poetry.lock
forge:
- classic/forge/forge/**
- classic/forge/tests/**
- classic/forge/poetry.lock
benchmark:
- classic/benchmark/agbenchmark/**
- classic/benchmark/tests/**
- classic/benchmark/poetry.lock
outputs:
changed-parts: ${{ steps.changes-in.outputs.changes }}
lint:
needs: get-changed-parts
runs-on: ubuntu-latest
env:
min-python-version: "3.10"
strategy:
matrix:
sub-package: ${{ fromJson(needs.get-changed-parts.outputs.changed-parts) }}
fail-fast: false
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Set up Python ${{ env.min-python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ env.min-python-version }}
- name: Set up Python dependency cache
uses: actions/cache@v4
with:
path: ~/.cache/pypoetry
key: ${{ runner.os }}-poetry-${{ hashFiles(format('{0}/poetry.lock', matrix.sub-package)) }}
- name: Install Poetry
run: curl -sSL https://install.python-poetry.org | python3 -
# Install dependencies
- name: Install Python dependencies
run: poetry -C classic/${{ matrix.sub-package }} install
# Lint
- name: Lint (isort)
run: poetry run isort --check .
working-directory: classic/${{ matrix.sub-package }}
- name: Lint (Black)
if: success() || failure()
run: poetry run black --check .
working-directory: classic/${{ matrix.sub-package }}
- name: Lint (Flake8)
if: success() || failure()
run: poetry run flake8 .
working-directory: classic/${{ matrix.sub-package }}
types:
needs: get-changed-parts
runs-on: ubuntu-latest
env:
min-python-version: "3.10"
strategy:
matrix:
sub-package: ${{ fromJson(needs.get-changed-parts.outputs.changed-parts) }}
fail-fast: false
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Set up Python ${{ env.min-python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ env.min-python-version }}
- name: Set up Python dependency cache
uses: actions/cache@v4
with:
path: ~/.cache/pypoetry
key: ${{ runner.os }}-poetry-${{ hashFiles(format('{0}/poetry.lock', matrix.sub-package)) }}
- name: Install Poetry
run: curl -sSL https://install.python-poetry.org | python3 -
# Install dependencies
- name: Install Python dependencies
run: poetry -C classic/${{ matrix.sub-package }} install
# Typecheck
- name: Typecheck
if: success() || failure()
run: poetry run pyright
working-directory: classic/${{ matrix.sub-package }}

View File

@@ -1,11 +1,11 @@
name: Classic - Purge Auto-GPT Docker CI cache
name: Purge Docker CI cache
on:
schedule:
- cron: 20 4 * * 1,4
env:
BASE_BRANCH: development
BASE_BRANCH: master
IMAGE_NAME: auto-gpt
jobs:
@@ -16,21 +16,19 @@ jobs:
build-type: [release, dev]
steps:
- name: Checkout repository
uses: actions/checkout@v4
uses: actions/checkout@v3
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
uses: docker/setup-buildx-action@v2
- id: build
name: Build image
uses: docker/build-push-action@v5
uses: docker/build-push-action@v3
with:
context: classic/
file: classic/Dockerfile.autogpt
build-args: BUILD_TYPE=${{ matrix.build-type }}
load: true # save to docker images
# use GHA cache as read-only
cache-to: type=gha,scope=autogpt-docker-${{ matrix.build-type }},mode=max
cache-to: type=gha,scope=docker-${{ matrix.build-type }},mode=max
- name: Generate build report
env:
@@ -39,10 +37,10 @@ jobs:
build_type: ${{ matrix.build-type }}
prod_branch: master
dev_branch: development
prod_branch: stable
dev_branch: master
repository: ${{ github.repository }}
base_branch: ${{ github.ref_name != 'master' && github.ref_name != 'development' && 'development' || 'master' }}
base_branch: ${{ github.ref_name != 'master' && github.ref_name != 'stable' && 'master' || 'stable' }}
current_ref: ${{ github.ref_name }}
commit_hash: ${{ github.sha }}

124
.github/workflows/docker-ci.yml vendored Normal file
View File

@@ -0,0 +1,124 @@
name: Docker CI
on:
push:
branches: [ master ]
paths-ignore:
- 'tests/Auto-GPT-test-cassettes'
- 'tests/challenges/current_score.json'
pull_request:
branches: [ master, release-*, stable ]
concurrency:
group: ${{ format('docker-ci-{0}', github.head_ref && format('pr-{0}', github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ github.event_name == 'pull_request' }}
env:
IMAGE_NAME: auto-gpt
jobs:
build:
runs-on: ubuntu-latest
strategy:
matrix:
build-type: [release, dev]
steps:
- name: Checkout repository
uses: actions/checkout@v3
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v2
- if: runner.debug
run: |
ls -al
du -hs *
- id: build
name: Build image
uses: docker/build-push-action@v3
with:
build-args: BUILD_TYPE=${{ matrix.build-type }}
tags: ${{ env.IMAGE_NAME }}
load: true # save to docker images
# cache layers in GitHub Actions cache to speed up builds
cache-from: type=gha,scope=docker-${{ matrix.build-type }}
cache-to: type=gha,scope=docker-${{ matrix.build-type }},mode=max
- name: Generate build report
env:
event_name: ${{ github.event_name }}
event_ref: ${{ github.event.ref }}
event_ref_type: ${{ github.event.ref}}
build_type: ${{ matrix.build-type }}
prod_branch: stable
dev_branch: master
repository: ${{ github.repository }}
base_branch: ${{ github.ref_name != 'master' && github.ref_name != 'stable' && 'master' || 'stable' }}
current_ref: ${{ github.ref_name }}
commit_hash: ${{ github.event.after }}
source_url: ${{ format('{0}/tree/{1}', github.event.repository.url, github.event.release && github.event.release.tag_name || github.sha) }}
push_forced_label: ${{ github.event.forced && '☢️ forced' || '' }}
new_commits_json: ${{ toJSON(github.event.commits) }}
compare_url_template: ${{ format('/{0}/compare/{{base}}...{{head}}', github.repository) }}
github_context_json: ${{ toJSON(github) }}
job_env_json: ${{ toJSON(env) }}
vars_json: ${{ toJSON(vars) }}
run: .github/workflows/scripts/docker-ci-summary.sh >> $GITHUB_STEP_SUMMARY
continue-on-error: true
test:
runs-on: ubuntu-latest
timeout-minutes: 10
steps:
- name: Check out repository
uses: actions/checkout@v3
with:
submodules: true
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v2
- id: build
name: Build image
uses: docker/build-push-action@v3
with:
build-args: BUILD_TYPE=dev # include pytest
tags: ${{ env.IMAGE_NAME }}
load: true # save to docker images
# cache layers in GitHub Actions cache to speed up builds
cache-from: type=gha,scope=docker-dev
cache-to: type=gha,scope=docker-dev,mode=max
- id: test
name: Run tests
env:
CI: true
PLAIN_OUTPUT: True
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
run: |
set +e
test_output=$(
docker run --env CI --env OPENAI_API_KEY --entrypoint python ${{ env.IMAGE_NAME }} -m \
pytest -v --cov=autogpt --cov-branch --cov-report term-missing \
--numprocesses=4 --durations=10 \
tests/unit tests/integration 2>&1
)
test_failure=$?
echo "$test_output"
cat << $EOF >> $GITHUB_STEP_SUMMARY
# Tests $([ $test_failure = 0 ] && echo '✅' || echo '❌')
\`\`\`
$test_output
\`\`\`
$EOF
exit $test_failure

View File

@@ -1,4 +1,4 @@
name: Classic - AutoGPT Docker Release
name: Docker Release
on:
release:
@@ -16,36 +16,31 @@ env:
jobs:
build:
if: startsWith(github.ref, 'refs/tags/autogpt-')
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v4
uses: actions/checkout@v3
- name: Log in to Docker hub
uses: docker/login-action@v3
uses: docker/login-action@v2
with:
username: ${{ secrets.DOCKER_USER }}
password: ${{ secrets.DOCKER_PASSWORD }}
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
uses: docker/setup-buildx-action@v2
# slashes are not allowed in image tags, but can appear in git branch or tag names
- id: sanitize_tag
name: Sanitize image tag
run: |
tag=${raw_tag//\//-}
echo tag=${tag#autogpt-} >> $GITHUB_OUTPUT
run: echo tag=${raw_tag//\//-} >> $GITHUB_OUTPUT
env:
raw_tag: ${{ github.ref_name }}
- id: build
name: Build image
uses: docker/build-push-action@v5
uses: docker/build-push-action@v3
with:
context: classic/
file: Dockerfile.autogpt
build-args: BUILD_TYPE=release
load: true # save to docker images
# push: true # TODO: uncomment when this issue is fixed: https://github.com/moby/buildkit/issues/1555
@@ -53,11 +48,10 @@ jobs:
${{ env.IMAGE_NAME }},
${{ env.DEPLOY_IMAGE_NAME }}:latest,
${{ env.DEPLOY_IMAGE_NAME }}:${{ steps.sanitize_tag.outputs.tag }}
labels: GIT_REVISION=${{ github.sha }}
# cache layers in GitHub Actions cache to speed up builds
cache-from: ${{ !inputs.no_cache && 'type=gha' || '' }},scope=autogpt-docker-release
cache-to: type=gha,scope=autogpt-docker-release,mode=max
cache-from: ${{ !inputs.no_cache && 'type=gha' || '' }},scope=docker-release
cache-to: type=gha,scope=docker-release,mode=max
- name: Push image to Docker Hub
run: docker push --all-tags ${{ env.DEPLOY_IMAGE_NAME }}
@@ -69,10 +63,10 @@ jobs:
event_ref_type: ${{ github.event.ref}}
inputs_no_cache: ${{ inputs.no_cache }}
prod_branch: master
dev_branch: development
prod_branch: stable
dev_branch: master
repository: ${{ github.repository }}
base_branch: ${{ github.ref_name != 'master' && github.ref_name != 'development' && 'development' || 'master' }}
base_branch: ${{ github.ref_name != 'master' && github.ref_name != 'stable' && 'master' || 'stable' }}
ref_type: ${{ github.ref_type }}
current_ref: ${{ github.ref_name }}

View File

@@ -0,0 +1,37 @@
name: Docs
on:
push:
branches: [ stable ]
paths:
- 'docs/**'
- 'mkdocs.yml'
- '.github/workflows/documentation.yml'
# Allows you to run this workflow manually from the Actions tab
workflow_dispatch:
permissions:
contents: write
jobs:
deploy:
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v3
- name: Set up Python 3
uses: actions/setup-python@v4
with:
python-version: 3.x
- name: Set up workflow cache
uses: actions/cache@v3
with:
key: ${{ github.ref }}
path: .cache
- run: pip install mkdocs-material
- run: mkdocs gh-deploy --force

View File

@@ -1,41 +0,0 @@
name: Platform - AutoGPT Builder CI
on:
push:
branches: [ master ]
paths:
- '.github/workflows/autogpt-builder-ci.yml'
- 'autogpt_platform/autogpt_builder/**'
pull_request:
paths:
- '.github/workflows/autogpt-builder-ci.yml'
- 'autogpt_platform/autogpt_builder/**'
defaults:
run:
shell: bash
working-directory: autogpt_platform/autogpt_builder
jobs:
lint:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Set up Node.js
uses: actions/setup-node@v4
with:
node-version: '21'
- name: Install dependencies
run: |
npm install
- name: Check formatting with Prettier
run: |
npx prettier --check .
- name: Run lint
run: |
npm run lint

View File

@@ -1,56 +0,0 @@
name: Platform - AutoGPT Builder Infra
on:
push:
branches: [ master ]
paths:
- '.github/workflows/autogpt-infra-ci.yml'
- 'autogpt_platform/infra/**'
pull_request:
paths:
- '.github/workflows/autogpt-infra-ci.yml'
- 'autogpt_platform/infra/**'
defaults:
run:
shell: bash
working-directory: autogpt_platform/infra
jobs:
lint:
runs-on: ubuntu-latest
steps:
- name: Checkout
uses: actions/checkout@v2
with:
fetch-depth: 0
- name: TFLint
uses: pauloconnor/tflint-action@v0.0.2
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
tflint_path: terraform/
tflint_recurse: true
tflint_changed_only: false
- name: Set up Helm
uses: azure/setup-helm@v4.2.0
with:
version: v3.14.4
- name: Set up chart-testing
uses: helm/chart-testing-action@v2.6.0
- name: Run chart-testing (list-changed)
id: list-changed
run: |
changed=$(ct list-changed --target-branch ${{ github.event.repository.default_branch }})
if [[ -n "$changed" ]]; then
echo "changed=true" >> "$GITHUB_OUTPUT"
fi
- name: Run chart-testing (lint)
if: steps.list-changed.outputs.changed == 'true'
run: ct lint --target-branch ${{ github.event.repository.default_branch }}

View File

@@ -1,155 +0,0 @@
name: Platform - AutoGPT Server CI
on:
push:
branches: [master, development, ci-test*]
paths:
- ".github/workflows/autogpt-server-ci.yml"
- "autogpt_platform/autogpt_server/**"
pull_request:
branches: [master, development, release-*]
paths:
- ".github/workflows/autogpt-server-ci.yml"
- "autogpt_platform/autogpt_server/**"
concurrency:
group: ${{ format('autogpt-server-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
cancel-in-progress: ${{ startsWith(github.event_name, 'pull_request') }}
defaults:
run:
shell: bash
working-directory: autogpt_platform/autogpt_server
jobs:
test:
permissions:
contents: read
timeout-minutes: 30
strategy:
fail-fast: false
matrix:
python-version: ["3.10"]
platform-os: [ubuntu, macos, macos-arm64, windows]
runs-on: ${{ matrix.platform-os != 'macos-arm64' && format('{0}-latest', matrix.platform-os) || 'macos-14' }}
steps:
- name: Setup PostgreSQL
uses: ikalnytskyi/action-setup-postgres@v6
with:
username: ${{ secrets.DB_USER || 'postgres' }}
password: ${{ secrets.DB_PASS || 'postgres' }}
database: postgres
port: 5432
id: postgres
# Quite slow on macOS (2~4 minutes to set up Docker)
# - name: Set up Docker (macOS)
# if: runner.os == 'macOS'
# uses: crazy-max/ghaction-setup-docker@v3
- name: Start MinIO service (Linux)
if: runner.os == 'Linux'
working-directory: "."
run: |
docker pull minio/minio:edge-cicd
docker run -d -p 9000:9000 minio/minio:edge-cicd
- name: Start MinIO service (macOS)
if: runner.os == 'macOS'
working-directory: ${{ runner.temp }}
run: |
brew install minio/stable/minio
mkdir data
minio server ./data &
# No MinIO on Windows:
# - Windows doesn't support running Linux Docker containers
# - It doesn't seem possible to start background processes on Windows. They are
# killed after the step returns.
# See: https://github.com/actions/runner/issues/598#issuecomment-2011890429
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
submodules: true
- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
- id: get_date
name: Get date
run: echo "date=$(date +'%Y-%m-%d')" >> $GITHUB_OUTPUT
- name: Set up Python dependency cache
# On Windows, unpacking cached dependencies takes longer than just installing them
if: runner.os != 'Windows'
uses: actions/cache@v4
with:
path: ${{ runner.os == 'macOS' && '~/Library/Caches/pypoetry' || '~/.cache/pypoetry' }}
key: poetry-${{ runner.os }}-${{ hashFiles('autogpt_platform/autogpt_server/poetry.lock') }}
- name: Install Poetry (Unix)
if: runner.os != 'Windows'
run: |
curl -sSL https://install.python-poetry.org | python3 -
if [ "${{ runner.os }}" = "macOS" ]; then
PATH="$HOME/.local/bin:$PATH"
echo "$HOME/.local/bin" >> $GITHUB_PATH
fi
- name: Install Poetry (Windows)
if: runner.os == 'Windows'
shell: pwsh
run: |
(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -
$env:PATH += ";$env:APPDATA\Python\Scripts"
echo "$env:APPDATA\Python\Scripts" >> $env:GITHUB_PATH
- name: Install Python dependencies
run: poetry install
- name: Generate Prisma Client
run: poetry run prisma generate
- name: Run Database Migrations
run: poetry run prisma migrate dev --name updates
env:
CONNECTION_STR: ${{ steps.postgres.outputs.connection-uri }}
- id: lint
name: Run Linter
run: poetry run lint
- name: Run pytest with coverage
run: |
if [[ "${{ runner.debug }}" == "1" ]]; then
poetry run pytest -vv -o log_cli=true -o log_cli_level=DEBUG test
else
poetry run pytest -vv test
fi
if: success() || (failure() && steps.lint.outcome == 'failure')
env:
LOG_LEVEL: ${{ runner.debug && 'DEBUG' || 'INFO' }}
env:
CI: true
PLAIN_OUTPUT: True
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
DB_USER: ${{ secrets.DB_USER || 'postgres' }}
DB_PASS: ${{ secrets.DB_PASS || 'postgres' }}
DB_NAME: postgres
DB_PORT: 5432
RUN_ENV: local
PORT: 8080
DATABASE_URL: postgresql://${{ secrets.DB_USER || 'postgres' }}:${{ secrets.DB_PASS || 'postgres' }}@localhost:5432/${{ secrets.DB_NAME || 'postgres'}}
# - name: Upload coverage reports to Codecov
# uses: codecov/codecov-action@v4
# with:
# token: ${{ secrets.CODECOV_TOKEN }}
# flags: autogpt-server,${{ runner.os }}

View File

@@ -1,12 +1,12 @@
name: Repo - Pull Request auto-label
name: "Pull Request auto-label"
on:
# So that PRs touching the same files as the push are updated
push:
branches: [ master, development, release-* ]
branches: [ master, release-* ]
paths-ignore:
- 'classic/forge/tests/vcr_cassettes'
- 'classic/benchmark/reports/**'
- 'tests/Auto-GPT-test-cassettes'
- 'tests/challenges/current_score.json'
# So that the `dirtyLabel` is removed if conflicts are resolve
# We recommend `pull_request_target` so that github secrets are available.
# In `pull_request` we wouldn't be able to change labels of fork PRs
@@ -52,15 +52,6 @@ jobs:
l_label: 'size/l'
l_max_size: 500
xl_label: 'size/xl'
message_if_xl:
scope:
if: ${{ github.event_name == 'pull_request_target' }}
permissions:
contents: read
pull-requests: write
runs-on: ubuntu-latest
steps:
- uses: actions/labeler@v5
with:
sync-labels: true
message_if_xl: >
This PR exceeds the recommended size of 500 lines.
Please make sure you are NOT addressing multiple issues with one PR.

View File

@@ -1,34 +0,0 @@
name: Repo - Close stale issues
on:
schedule:
- cron: '30 1 * * *'
workflow_dispatch:
permissions:
issues: write
jobs:
stale:
runs-on: ubuntu-latest
steps:
- uses: actions/stale@v9
with:
# operations-per-run: 5000
stale-issue-message: >
This issue has automatically been marked as _stale_ because it has not had
any activity in the last 50 days. You can _unstale_ it by commenting or
removing the label. Otherwise, this issue will be closed in 10 days.
stale-pr-message: >
This pull request has automatically been marked as _stale_ because it has
not had any activity in the last 50 days. You can _unstale_ it by commenting
or removing the label.
close-issue-message: >
This issue was closed automatically because it has been stale for 10 days
with no activity.
days-before-stale: 50
days-before-close: 10
# Do not touch meta issues:
exempt-issue-labels: meta,fridge,project management
# Do not affect pull requests:
days-before-pr-stale: -1
days-before-pr-close: -1

View File

@@ -1,20 +0,0 @@
name: Repo - Github Stats
on:
schedule:
# Run this once per day, towards the end of the day for keeping the most
# recent data point most meaningful (hours are interpreted in UTC).
- cron: "0 23 * * *"
workflow_dispatch: # Allow for running this manually.
jobs:
j1:
name: github-repo-stats
runs-on: ubuntu-latest
steps:
- name: run-ghrs
# Use latest release.
uses: jgehrcke/github-repo-stats@HEAD
with:
ghtoken: ${{ secrets.ghrs_github_api_token }}

View File

@@ -1,31 +0,0 @@
name: Repo - PR Status Checker
on:
pull_request:
types: [opened, synchronize, reopened]
jobs:
status-check:
name: Check PR Status
runs-on: ubuntu-latest
steps:
# - name: Wait some time for all actions to start
# run: sleep 30
- uses: actions/checkout@v4
# with:
# fetch-depth: 0
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: "3.10"
- name: Install dependencies
run: |
python -m pip install --upgrade pip
pip install requests
- name: Check PR Status
run: |
echo "Current directory before running Python script:"
pwd
echo "Attempting to run Python script:"
python .github/workflows/scripts/check_actions_status.py
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

View File

@@ -1,109 +0,0 @@
import json
import os
import requests
import sys
import time
from typing import Dict, List, Tuple
def get_environment_variables() -> Tuple[str, str, str, str, str]:
"""Retrieve and return necessary environment variables."""
try:
with open(os.environ["GITHUB_EVENT_PATH"]) as f:
event = json.load(f)
sha = event["pull_request"]["head"]["sha"]
return (
os.environ["GITHUB_API_URL"],
os.environ["GITHUB_REPOSITORY"],
sha,
os.environ["GITHUB_TOKEN"],
os.environ["GITHUB_RUN_ID"],
)
except KeyError as e:
print(f"Error: Missing required environment variable or event data: {e}")
sys.exit(1)
def make_api_request(url: str, headers: Dict[str, str]) -> Dict:
"""Make an API request and return the JSON response."""
try:
print("Making API request to:", url)
response = requests.get(url, headers=headers, timeout=10)
response.raise_for_status()
return response.json()
except requests.RequestException as e:
print(f"Error: API request failed. {e}")
sys.exit(1)
def process_check_runs(check_runs: List[Dict]) -> Tuple[bool, bool]:
"""Process check runs and return their status."""
runs_in_progress = False
all_others_passed = True
for run in check_runs:
if str(run["name"]) != "Check PR Status":
status = run["status"]
conclusion = run["conclusion"]
if status == "completed":
if conclusion not in ["success", "skipped", "neutral"]:
all_others_passed = False
print(
f"Check run {run['name']} (ID: {run['id']}) has conclusion: {conclusion}"
)
else:
runs_in_progress = True
print(f"Check run {run['name']} (ID: {run['id']}) is still {status}.")
all_others_passed = False
else:
print(
f"Skipping check run {run['name']} (ID: {run['id']}) as it is the current run."
)
return runs_in_progress, all_others_passed
def main():
api_url, repo, sha, github_token, current_run_id = get_environment_variables()
endpoint = f"{api_url}/repos/{repo}/commits/{sha}/check-runs"
headers = {
"Accept": "application/vnd.github.v3+json",
}
if github_token:
headers["Authorization"] = f"token {github_token}"
print(f"Current run ID: {current_run_id}")
while True:
data = make_api_request(endpoint, headers)
check_runs = data["check_runs"]
print("Processing check runs...")
print(check_runs)
runs_in_progress, all_others_passed = process_check_runs(check_runs)
if not runs_in_progress:
break
print(
"Some check runs are still in progress. Waiting 3 minutes before checking again..."
)
time.sleep(180)
if all_others_passed:
print("All other completed check runs have passed. This check passes.")
sys.exit(0)
else:
print("Some check runs have failed or have not completed. This check fails.")
sys.exit(1)
if __name__ == "__main__":
main()

27
.gitignore vendored
View File

@@ -1,11 +1,12 @@
## Original ignores
.github_access_token
classic/original_autogpt/keys.py
classic/original_autogpt/*.json
autogpt/keys.py
autogpt/*.json
auto_gpt_workspace/*
*.mpeg
.env
azure.yaml
ai_settings.yaml
last_run_ai_settings.yaml
.vscode
.idea/*
auto-gpt.json
@@ -27,11 +28,15 @@ __pycache__/
# Distribution / packaging
.Python
build/
develop-eggs/
dist/
/plugins/
plugins_config.yaml
downloads/
eggs/
.eggs/
lib/
lib64/
parts/
sdist/
@@ -155,19 +160,3 @@ openai/
# news
CURRENT_BULLETIN.md
# AgBenchmark
agclassic/benchmark/reports/
# Nodejs
package-lock.json
# Allow for locally private items
# private
pri*
# ignore
ig*
.github_access_token
LICENSE.rtf
autogpt_platform/autogpt_server/settings.py

10
.gitmodules vendored
View File

@@ -1,6 +1,4 @@
[submodule "classic/forge/tests/vcr_cassettes"]
path = classic/forge/tests/vcr_cassettes
url = https://github.com/Significant-Gravitas/Auto-GPT-test-cassettes
[submodule "autogpt_platform/supabase"]
path = autogpt_platform/supabase
url = https://github.com/supabase/supabase.git
[submodule "tests/Auto-GPT-test-cassettes"]
path = tests/Auto-GPT-test-cassettes
url = https://github.com/Significant-Gravitas/Auto-GPT-test-cassettes
branch = master

10
.isort.cfg Normal file
View File

@@ -0,0 +1,10 @@
[settings]
profile = black
multi_line_output = 3
include_trailing_comma = true
force_grid_wrap = 0
use_parentheses = true
ensure_newline_before_comments = true
line_length = 88
sections = FUTURE,STDLIB,THIRDPARTY,FIRSTPARTY,LOCALFOLDER
skip = .tox,__pycache__,*.pyc,venv*/*,reports,venv,env,node_modules,.env,.venv,dist

View File

@@ -1,6 +0,0 @@
[pr_reviewer]
num_code_suggestions=0
[pr_code_suggestions]
commitable_code_suggestions=false
num_code_suggestions=0

View File

@@ -3,125 +3,40 @@ repos:
rev: v4.4.0
hooks:
- id: check-added-large-files
args: ["--maxkb=500"]
- id: fix-byte-order-marker
args: ['--maxkb=500']
- id: check-byte-order-marker
- id: check-case-conflict
- id: check-merge-conflict
- id: check-symlinks
- id: debug-statements
- repo: local
# isort needs the context of which packages are installed to function, so we
# can't use a vendored isort pre-commit hook (which runs in its own isolated venv).
- repo: https://github.com/pycqa/isort
rev: 5.12.0
hooks:
- id: isort-autogpt
name: Lint (isort) - AutoGPT
entry: poetry -C classic/original_autogpt run isort
files: ^classic/original_autogpt/
types: [file, python]
language: system
- id: isort-forge
name: Lint (isort) - Forge
entry: poetry -C classic/forge run isort
files: ^classic/forge/
types: [file, python]
language: system
- id: isort-benchmark
name: Lint (isort) - Benchmark
entry: poetry -C classic/benchmark run isort
files: ^classic/benchmark/
types: [file, python]
language: system
- repo: https://github.com/psf/black
rev: 23.12.1
# Black has sensible defaults, doesn't need package context, and ignores
# everything in .gitignore, so it works fine without any config or arguments.
hooks:
- id: black
name: Lint (Black)
- id: isort
language_version: python3.10
- repo: https://github.com/PyCQA/flake8
rev: 7.0.0
# To have flake8 load the config of the individual subprojects, we have to call
# them separately.
- repo: https://github.com/psf/black
rev: 23.3.0
hooks:
- id: flake8
name: Lint (Flake8) - AutoGPT
alias: flake8-autogpt
files: ^classic/original_autogpt/(autogpt|scripts|tests)/
args: [--config=classic/original_autogpt/.flake8]
- id: black
language_version: python3.10
- id: flake8
name: Lint (Flake8) - Forge
alias: flake8-forge
files: ^classic/forge/(forge|tests)/
args: [--config=classic/forge/.flake8]
- id: flake8
name: Lint (Flake8) - Benchmark
alias: flake8-benchmark
files: ^classic/benchmark/(agbenchmark|tests)/((?!reports).)*[/.]
args: [--config=classic/benchmark/.flake8]
- repo: local
# To have watertight type checking, we check *all* the files in an affected
# project. To trigger on poetry.lock we also reset the file `types` filter.
- repo: https://github.com/pre-commit/mirrors-mypy
rev: 'v1.3.0'
hooks:
- id: pyright
name: Typecheck - AutoGPT
alias: pyright-autogpt
entry: poetry -C classic/original_autogpt run pyright
args: [-p, autogpt, autogpt]
# include forge source (since it's a path dependency) but exclude *_test.py files:
files: ^(classic/original_autogpt/((autogpt|scripts|tests)/|poetry\.lock$)|classic/forge/(classic/forge/.*(?<!_test)\.py|poetry\.lock)$)
types: [file]
language: system
pass_filenames: false
- id: pyright
name: Typecheck - Forge
alias: pyright-forge
entry: poetry -C classic/forge run pyright
args: [-p, forge, forge]
files: ^classic/forge/(classic/forge/|poetry\.lock$)
types: [file]
language: system
pass_filenames: false
- id: pyright
name: Typecheck - Benchmark
alias: pyright-benchmark
entry: poetry -C classic/benchmark run pyright
args: [-p, benchmark, benchmark]
files: ^classic/benchmark/(agclassic/benchmark/|tests/|poetry\.lock$)
types: [file]
language: system
pass_filenames: false
- id: mypy
- repo: local
hooks:
- id: pytest-autogpt
name: Run tests - AutoGPT (excl. slow tests)
entry: bash -c 'cd classic/original_autogpt && poetry run pytest --cov=autogpt -m "not slow" tests/unit tests/integration'
# include forge source (since it's a path dependency) but exclude *_test.py files:
files: ^(classic/original_autogpt/((autogpt|tests)/|poetry\.lock$)|classic/forge/(classic/forge/.*(?<!_test)\.py|poetry\.lock)$)
language: system
pass_filenames: false
- id: pytest-forge
name: Run tests - Forge (excl. slow tests)
entry: bash -c 'cd classic/forge && poetry run pytest --cov=forge -m "not slow"'
files: ^classic/forge/(classic/forge/|tests/|poetry\.lock$)
language: system
pass_filenames: false
- id: pytest-benchmark
name: Run tests - Benchmark
entry: bash -c 'cd classic/benchmark && poetry run pytest --cov=benchmark'
files: ^classic/benchmark/(agclassic/benchmark/|tests/|poetry\.lock$)
- id: autoflake
name: autoflake
entry: autoflake --in-place --remove-all-unused-imports --recursive --ignore-init-module-imports --ignore-pass-after-docstring autogpt tests
language: python
types: [ python ]
- id: pytest-check
name: pytest-check
entry: pytest --cov=autogpt tests/unit
language: system
pass_filenames: false
always_run: true

View File

@@ -1,61 +0,0 @@
{
"folders": [
{
"name": "autogpt_server",
"path": "../autogpt_platform/autogpt_server"
},
{
"name": "autogpt_builder",
"path": "../autogpt_platform/autogpt_builder"
},
{
"name": "market",
"path": "../autogpt_platform/market"
},
{
"name": "lib",
"path": "../autogpt_platform/autogpt_libs"
},
{
"name": "infra",
"path": "../autogpt_platform/infra"
},
{
"name": "docs",
"path": "../docs"
},
{
"name": "[root]",
"path": ".."
},
{
"name": "classic - autogpt",
"path": "../classic/original_autogpt"
},
{
"name": "classic - benchmark",
"path": "../classic/benchmark"
},
{
"name": "classic - forge",
"path": "../classic/forge"
},
{
"name": "classic - frontend",
"path": "../classic/frontend"
},
],
"settings": {
"python.analysis.typeCheckingMode": "basic"
},
"extensions": {
"recommendations": [
"charliermarsh.ruff",
"dart-code.flutter",
"ms-python.black-formatter",
"ms-python.vscode-pylance",
"prisma.prisma",
"qwtel.sqlite-viewer"
]
}
}

27
BULLETIN.md Normal file
View File

@@ -0,0 +1,27 @@
# QUICK LINKS 🔗
# --------------
🌎 *Official Website*: https://agpt.co.
📖 *User Guide*: https://docs.agpt.co.
👩 *Contributors Wiki*: https://github.com/Significant-Gravitas/Auto-GPT/wiki/Contributing.
# v0.4.6 RELEASE HIGHLIGHTS! 🚀
# -----------------------------
This release includes under-the-hood improvements and bug fixes, including better UTF-8
special character support, workspace write access for sandboxed Python execution,
more robust path resolution for config files and the workspace, and a full restructure
of the Agent class, the "brain" of Auto-GPT, to make it more extensible.
We have also released some documentation updates, including:
- *How to share your system logs*
Visit [docs/share-your-logs.md] to learn how to how to share logs with us
via a log analyzer graciously contributed by https://www.e2b.dev/
- *Auto-GPT re-architecture documentation*
You can learn more about the inner-workings of the Auto-GPT re-architecture
released last cycle, via these links:
* [autogpt/core/README.md]
* [autogpt/core/ARCHITECTURE_NOTES.md]
Take a look at the Release Notes on Github for the full changelog!
https://github.com/Significant-Gravitas/Auto-GPT/releases.

View File

@@ -1,21 +0,0 @@
# This CITATION.cff file was generated with cffinit.
# Visit https://bit.ly/cffinit to generate yours today!
cff-version: 1.2.0
title: AutoGPT
message: >-
If you use this software, please cite it using the
metadata from this file.
type: software
authors:
- name: Significant Gravitas
website: 'https://agpt.co'
repository-code: 'https://github.com/Significant-Gravitas/AutoGPT'
url: 'https://agpt.co'
abstract: >-
A collection of tools and experimental open-source attempts to make GPT-4 fully
autonomous.
keywords:
- AI
- Agent
license: MIT

View File

@@ -1,12 +1,12 @@
# Code of Conduct for AutoGPT
# Code of Conduct for Auto-GPT
## 1. Purpose
The purpose of this Code of Conduct is to provide guidelines for contributors to the AutoGPT projects on GitHub. We aim to create a positive and inclusive environment where all participants can contribute and collaborate effectively. By participating in this project, you agree to abide by this Code of Conduct.
The purpose of this Code of Conduct is to provide guidelines for contributors to the auto-gpt project on GitHub. We aim to create a positive and inclusive environment where all participants can contribute and collaborate effectively. By participating in this project, you agree to abide by this Code of Conduct.
## 2. Scope
This Code of Conduct applies to all contributors, maintainers, and users of the AutoGPT project. It extends to all project spaces, including but not limited to issues, pull requests, code reviews, comments, and other forms of communication within the project.
This Code of Conduct applies to all contributors, maintainers, and users of the auto-gpt project. It extends to all project spaces, including but not limited to issues, pull requests, code reviews, comments, and other forms of communication within the project.
## 3. Our Standards
@@ -36,5 +36,4 @@ This Code of Conduct is adapted from the [Contributor Covenant](https://www.cont
## 6. Contact
If you have any questions or concerns, please contact the project maintainers on Discord:
https://discord.gg/autogpt
If you have any questions or concerns, please contact the project maintainers.

View File

@@ -1,38 +1,14 @@
# AutoGPT Contribution Guide
If you are reading this, you are probably looking for the full **[contribution guide]**,
which is part of our [wiki].
We maintain a knowledgebase at this [wiki](https://github.com/Significant-Gravitas/Nexus/wiki)
Also check out our [🚀 Roadmap][roadmap] for information about our priorities and associated tasks.
<!-- You can find our immediate priorities and their progress on our public [kanban board]. -->
We would like to say "We value all contributions". After all, we are an open-source project, so we should say something fluffy like this, right?
[contribution guide]: https://github.com/Significant-Gravitas/AutoGPT/wiki/Contributing
[wiki]: https://github.com/Significant-Gravitas/AutoGPT/wiki
[roadmap]: https://github.com/Significant-Gravitas/AutoGPT/discussions/6971
[kanban board]: https://github.com/orgs/Significant-Gravitas/projects/1
However the reality is that some contributions are SUPER-valuable, while others create more trouble than they are worth and actually _create_ work for the core team.
## In short
1. Avoid duplicate work, issues, PRs etc.
2. We encourage you to collaborate with fellow community members on some of our bigger
[todo's][roadmap]!
* We highly recommend to post your idea and discuss it in the [dev channel].
3. Create a draft PR when starting work on bigger changes.
4. Adhere to the [Code Guidelines]
5. Clearly explain your changes when submitting a PR.
6. Don't submit broken code: test/validate your changes.
7. Avoid making unnecessary changes, especially if they're purely based on your personal
preferences. Doing so is the maintainers' job. ;-)
8. Please also consider contributing something other than code; see the
[contribution guide] for options.
If you wish to contribute, please look through the wiki [contributing](https://github.com/Significant-Gravitas/Nexus/wiki/Contributing) page.
[dev channel]: https://discord.com/channels/1092243196446249134/1095817829405704305
[code guidelines]: https://github.com/Significant-Gravitas/AutoGPT/wiki/Contributing#code-guidelines
If you wish to involve with the project (beyond just contributing PRs), please read the wiki [catalyzing](https://github.com/Significant-Gravitas/Nexus/wiki/Catalyzing) page.
If you wish to involve with the project (beyond just contributing PRs), please read the
wiki page about [Catalyzing](https://github.com/Significant-Gravitas/AutoGPT/wiki/Catalyzing).
In fact, why not just look through the whole wiki (it's only a few pages) and
hop on our Discord. See you there! :-)
In fact, why not just look through the whole wiki (it's only a few pages) and hop on our discord (you'll find it in the wiki).
❤️ & 🔆
The team @ AutoGPT
https://discord.gg/autogpt
The team @ Auto-GPT

46
Dockerfile Normal file
View File

@@ -0,0 +1,46 @@
# 'dev' or 'release' container build
ARG BUILD_TYPE=dev
# Use an official Python base image from the Docker Hub
FROM python:3.10-slim AS autogpt-base
# Install browsers
RUN apt-get update && apt-get install -y \
chromium-driver firefox-esr ca-certificates \
&& apt-get clean && rm -rf /var/lib/apt/lists/*
# Install utilities
RUN apt-get update && apt-get install -y \
curl jq wget git \
&& apt-get clean && rm -rf /var/lib/apt/lists/*
# Set environment variables
ENV PIP_NO_CACHE_DIR=yes \
PYTHONUNBUFFERED=1 \
PYTHONDONTWRITEBYTECODE=1
# Install the required python packages globally
ENV PATH="$PATH:/root/.local/bin"
COPY requirements.txt .
# Set the entrypoint
ENTRYPOINT ["python", "-m", "autogpt", "--install-plugin-deps"]
# dev build -> include everything
FROM autogpt-base as autogpt-dev
RUN pip install --no-cache-dir -r requirements.txt
WORKDIR /app
ONBUILD COPY . ./
# release build -> include bare minimum
FROM autogpt-base as autogpt-release
RUN sed -i '/Items below this point will not be included in the Docker Image/,$d' requirements.txt && \
pip install --no-cache-dir -r requirements.txt
WORKDIR /app
ONBUILD COPY autogpt/ ./autogpt
ONBUILD COPY scripts/ ./scripts
ONBUILD COPY plugins/ ./plugins
ONBUILD COPY prompt_settings.yaml ./prompt_settings.yaml
ONBUILD RUN mkdir ./data
FROM autogpt-${BUILD_TYPE} AS auto-gpt

233
README.md

File diff suppressed because one or more lines are too long

Binary file not shown.

Binary file not shown.

Before

Width:  |  Height:  |  Size: 1.1 MiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 49 KiB

View File

@@ -2,6 +2,13 @@ import os
import random
import sys
from dotenv import load_dotenv
if "pytest" in sys.argv or "pytest" in sys.modules or os.getenv("CI"):
print("Setting random seed to 42")
random.seed(42)
# Load the users .env file into environment variables
load_dotenv(verbose=True, override=True)
del load_dotenv

5
autogpt/__main__.py Normal file
View File

@@ -0,0 +1,5 @@
"""Auto-GPT: A GPT powered AI Assistant"""
import autogpt.app.cli
if __name__ == "__main__":
autogpt.app.cli.main()

View File

@@ -0,0 +1,4 @@
from .agent import Agent
from .base import AgentThoughts, BaseAgent, CommandArgs, CommandName
__all__ = ["BaseAgent", "Agent", "CommandName", "CommandArgs", "AgentThoughts"]

298
autogpt/agents/agent.py Normal file
View File

@@ -0,0 +1,298 @@
from __future__ import annotations
import json
import time
from datetime import datetime
from typing import TYPE_CHECKING, Any, Optional
if TYPE_CHECKING:
from autogpt.config import AIConfig, Config
from autogpt.llm.base import ChatModelResponse, ChatSequence
from autogpt.memory.vector import VectorMemory
from autogpt.models.command_registry import CommandRegistry
from autogpt.json_utils.utilities import extract_dict_from_response, validate_dict
from autogpt.llm.api_manager import ApiManager
from autogpt.llm.base import Message
from autogpt.llm.utils import count_string_tokens
from autogpt.logs import logger
from autogpt.logs.log_cycle import (
FULL_MESSAGE_HISTORY_FILE_NAME,
NEXT_ACTION_FILE_NAME,
USER_INPUT_FILE_NAME,
LogCycleHandler,
)
from autogpt.workspace import Workspace
from .base import AgentThoughts, BaseAgent, CommandArgs, CommandName
class Agent(BaseAgent):
"""Agent class for interacting with Auto-GPT."""
def __init__(
self,
ai_config: AIConfig,
command_registry: CommandRegistry,
memory: VectorMemory,
triggering_prompt: str,
config: Config,
cycle_budget: Optional[int] = None,
):
super().__init__(
ai_config=ai_config,
command_registry=command_registry,
config=config,
default_cycle_instruction=triggering_prompt,
cycle_budget=cycle_budget,
)
self.memory = memory
"""VectorMemoryProvider used to manage the agent's context (TODO)"""
self.workspace = Workspace(config.workspace_path, config.restrict_to_workspace)
"""Workspace that the agent has access to, e.g. for reading/writing files."""
self.created_at = datetime.now().strftime("%Y%m%d_%H%M%S")
"""Timestamp the agent was created; only used for structured debug logging."""
self.log_cycle_handler = LogCycleHandler()
"""LogCycleHandler for structured debug logging."""
def construct_base_prompt(self, *args, **kwargs) -> ChatSequence:
if kwargs.get("prepend_messages") is None:
kwargs["prepend_messages"] = []
# Clock
kwargs["prepend_messages"].append(
Message("system", f"The current time and date is {time.strftime('%c')}"),
)
# Add budget information (if any) to prompt
api_manager = ApiManager()
if api_manager.get_total_budget() > 0.0:
remaining_budget = (
api_manager.get_total_budget() - api_manager.get_total_cost()
)
if remaining_budget < 0:
remaining_budget = 0
budget_msg = Message(
"system",
f"Your remaining API budget is ${remaining_budget:.3f}"
+ (
" BUDGET EXCEEDED! SHUT DOWN!\n\n"
if remaining_budget == 0
else " Budget very nearly exceeded! Shut down gracefully!\n\n"
if remaining_budget < 0.005
else " Budget nearly exceeded. Finish up.\n\n"
if remaining_budget < 0.01
else ""
),
)
logger.debug(budget_msg)
if kwargs.get("append_messages") is None:
kwargs["append_messages"] = []
kwargs["append_messages"].append(budget_msg)
return super().construct_base_prompt(*args, **kwargs)
def on_before_think(self, *args, **kwargs) -> ChatSequence:
prompt = super().on_before_think(*args, **kwargs)
self.log_cycle_handler.log_count_within_cycle = 0
self.log_cycle_handler.log_cycle(
self.ai_config.ai_name,
self.created_at,
self.cycle_count,
self.history.raw(),
FULL_MESSAGE_HISTORY_FILE_NAME,
)
return prompt
def execute(
self,
command_name: str | None,
command_args: dict[str, str] | None,
user_input: str | None,
) -> str:
# Execute command
if command_name is not None and command_name.lower().startswith("error"):
result = f"Could not execute command: {command_name}{command_args}"
elif command_name == "human_feedback":
result = f"Human feedback: {user_input}"
self.log_cycle_handler.log_cycle(
self.ai_config.ai_name,
self.created_at,
self.cycle_count,
user_input,
USER_INPUT_FILE_NAME,
)
else:
for plugin in self.config.plugins:
if not plugin.can_handle_pre_command():
continue
command_name, arguments = plugin.pre_command(command_name, command_args)
command_result = execute_command(
command_name=command_name,
arguments=command_args,
agent=self,
)
result = f"Command {command_name} returned: " f"{command_result}"
result_tlength = count_string_tokens(str(command_result), self.llm.name)
memory_tlength = count_string_tokens(
str(self.history.summary_message()), self.llm.name
)
if result_tlength + memory_tlength > self.send_token_limit:
result = f"Failure: command {command_name} returned too much output. \
Do not execute this command again with the same arguments."
for plugin in self.config.plugins:
if not plugin.can_handle_post_command():
continue
result = plugin.post_command(command_name, result)
# Check if there's a result from the command append it to the message
if result is None:
self.history.add("system", "Unable to execute command", "action_result")
else:
self.history.add("system", result, "action_result")
return result
def parse_and_process_response(
self, llm_response: ChatModelResponse, *args, **kwargs
) -> tuple[CommandName | None, CommandArgs | None, AgentThoughts]:
if not llm_response.content:
raise SyntaxError("Assistant response has no text content")
assistant_reply_dict = extract_dict_from_response(llm_response.content)
valid, errors = validate_dict(assistant_reply_dict, self.config)
if not valid:
raise SyntaxError(
"Validation of response failed:\n "
+ ";\n ".join([str(e) for e in errors])
)
for plugin in self.config.plugins:
if not plugin.can_handle_post_planning():
continue
assistant_reply_dict = plugin.post_planning(assistant_reply_dict)
response = None, None, assistant_reply_dict
# Print Assistant thoughts
if assistant_reply_dict != {}:
# Get command name and arguments
try:
command_name, arguments = extract_command(
assistant_reply_dict, llm_response, self.config
)
response = command_name, arguments, assistant_reply_dict
except Exception as e:
logger.error("Error: \n", str(e))
self.log_cycle_handler.log_cycle(
self.ai_config.ai_name,
self.created_at,
self.cycle_count,
assistant_reply_dict,
NEXT_ACTION_FILE_NAME,
)
return response
def extract_command(
assistant_reply_json: dict, assistant_reply: ChatModelResponse, config: Config
) -> tuple[str, dict[str, str]]:
"""Parse the response and return the command name and arguments
Args:
assistant_reply_json (dict): The response object from the AI
assistant_reply (ChatModelResponse): The model response from the AI
config (Config): The config object
Returns:
tuple: The command name and arguments
Raises:
json.decoder.JSONDecodeError: If the response is not valid JSON
Exception: If any other error occurs
"""
if config.openai_functions:
if assistant_reply.function_call is None:
return "Error:", {"message": "No 'function_call' in assistant reply"}
assistant_reply_json["command"] = {
"name": assistant_reply.function_call.name,
"args": json.loads(assistant_reply.function_call.arguments),
}
try:
if "command" not in assistant_reply_json:
return "Error:", {"message": "Missing 'command' object in JSON"}
if not isinstance(assistant_reply_json, dict):
return (
"Error:",
{
"message": f"The previous message sent was not a dictionary {assistant_reply_json}"
},
)
command = assistant_reply_json["command"]
if not isinstance(command, dict):
return "Error:", {"message": "'command' object is not a dictionary"}
if "name" not in command:
return "Error:", {"message": "Missing 'name' field in 'command' object"}
command_name = command["name"]
# Use an empty dictionary if 'args' field is not present in 'command' object
arguments = command.get("args", {})
return command_name, arguments
except json.decoder.JSONDecodeError:
return "Error:", {"message": "Invalid JSON"}
# All other errors, return "Error: + error message"
except Exception as e:
return "Error:", {"message": str(e)}
def execute_command(
command_name: str,
arguments: dict[str, str],
agent: Agent,
) -> Any:
"""Execute the command and return the result
Args:
command_name (str): The name of the command to execute
arguments (dict): The arguments for the command
agent (Agent): The agent that is executing the command
Returns:
str: The result of the command
"""
try:
# Execute a native command with the same name or alias, if it exists
if command := agent.command_registry.get_command(command_name):
return command(**arguments, agent=agent)
# Handle non-native commands (e.g. from plugins)
for command in agent.ai_config.prompt_generator.commands:
if (
command_name == command["label"].lower()
or command_name == command["name"].lower()
):
return command["function"](**arguments)
raise RuntimeError(
f"Cannot execute '{command_name}': unknown command."
" Do not try to use this command again."
)
except Exception as e:
return f"Error: {str(e)}"

318
autogpt/agents/base.py Normal file
View File

@@ -0,0 +1,318 @@
from __future__ import annotations
from abc import ABCMeta, abstractmethod
from typing import TYPE_CHECKING, Any, Optional
if TYPE_CHECKING:
from autogpt.config import AIConfig, Config
from autogpt.models.command_registry import CommandRegistry
from autogpt.llm.base import ChatModelResponse, ChatSequence, Message
from autogpt.llm.providers.openai import OPEN_AI_CHAT_MODELS, get_openai_command_specs
from autogpt.llm.utils import count_message_tokens, create_chat_completion
from autogpt.logs import logger
from autogpt.memory.message_history import MessageHistory
from autogpt.prompts.prompt import DEFAULT_TRIGGERING_PROMPT
CommandName = str
CommandArgs = dict[str, str]
AgentThoughts = dict[str, Any]
class BaseAgent(metaclass=ABCMeta):
"""Base class for all Auto-GPT agents."""
def __init__(
self,
ai_config: AIConfig,
command_registry: CommandRegistry,
config: Config,
big_brain: bool = True,
default_cycle_instruction: str = DEFAULT_TRIGGERING_PROMPT,
cycle_budget: Optional[int] = 1,
send_token_limit: Optional[int] = None,
summary_max_tlength: Optional[int] = None,
):
self.ai_config = ai_config
"""The AIConfig or "personality" object associated with this agent."""
self.command_registry = command_registry
"""The registry containing all commands available to the agent."""
self.config = config
"""The applicable application configuration."""
self.big_brain = big_brain
"""
Whether this agent uses the configured smart LLM (default) to think,
as opposed to the configured fast LLM.
"""
self.default_cycle_instruction = default_cycle_instruction
"""The default instruction passed to the AI for a thinking cycle."""
self.cycle_budget = cycle_budget
"""
The number of cycles that the agent is allowed to run unsupervised.
`None` for unlimited continuous execution,
`1` to require user approval for every step,
`0` to stop the agent.
"""
self.cycles_remaining = cycle_budget
"""The number of cycles remaining within the `cycle_budget`."""
self.cycle_count = 0
"""The number of cycles that the agent has run since its initialization."""
self.system_prompt = ai_config.construct_full_prompt(config)
"""
The system prompt sets up the AI's personality and explains its goals,
available resources, and restrictions.
"""
llm_name = self.config.smart_llm if self.big_brain else self.config.fast_llm
self.llm = OPEN_AI_CHAT_MODELS[llm_name]
"""The LLM that the agent uses to think."""
self.send_token_limit = send_token_limit or self.llm.max_tokens * 3 // 4
"""
The token limit for prompt construction. Should leave room for the completion;
defaults to 75% of `llm.max_tokens`.
"""
self.history = MessageHistory(
self.llm,
max_summary_tlength=summary_max_tlength or self.send_token_limit // 6,
)
def think(
self,
instruction: Optional[str] = None,
) -> tuple[CommandName | None, CommandArgs | None, AgentThoughts]:
"""Runs the agent for one cycle.
Params:
instruction: The instruction to put at the end of the prompt.
Returns:
The command name and arguments, if any, and the agent's thoughts.
"""
instruction = instruction or self.default_cycle_instruction
prompt: ChatSequence = self.construct_prompt(instruction)
prompt = self.on_before_think(prompt, instruction)
raw_response = create_chat_completion(
prompt,
self.config,
functions=get_openai_command_specs(self.command_registry)
if self.config.openai_functions
else None,
)
self.cycle_count += 1
return self.on_response(raw_response, prompt, instruction)
@abstractmethod
def execute(
self,
command_name: str | None,
command_args: dict[str, str] | None,
user_input: str | None,
) -> str:
"""Executes the given command, if any, and returns the agent's response.
Params:
command_name: The name of the command to execute, if any.
command_args: The arguments to pass to the command, if any.
user_input: The user's input, if any.
Returns:
The results of the command.
"""
...
def construct_base_prompt(
self,
prepend_messages: list[Message] = [],
append_messages: list[Message] = [],
reserve_tokens: int = 0,
) -> ChatSequence:
"""Constructs and returns a prompt with the following structure:
1. System prompt
2. `prepend_messages`
3. Message history of the agent, truncated & prepended with running summary as needed
4. `append_messages`
Params:
prepend_messages: Messages to insert between the system prompt and message history
append_messages: Messages to insert after the message history
reserve_tokens: Number of tokens to reserve for content that is added later
"""
prompt = ChatSequence.for_model(
self.llm.name,
[Message("system", self.system_prompt)] + prepend_messages,
)
# Reserve tokens for messages to be appended later, if any
reserve_tokens += self.history.max_summary_tlength
if append_messages:
reserve_tokens += count_message_tokens(append_messages, self.llm.name)
# Fill message history, up to a margin of reserved_tokens.
# Trim remaining historical messages and add them to the running summary.
history_start_index = len(prompt)
trimmed_history = add_history_upto_token_limit(
prompt, self.history, self.send_token_limit - reserve_tokens
)
if trimmed_history:
new_summary_msg, _ = self.history.trim_messages(list(prompt), self.config)
prompt.insert(history_start_index, new_summary_msg)
if append_messages:
prompt.extend(append_messages)
return prompt
def construct_prompt(self, cycle_instruction: str) -> ChatSequence:
"""Constructs and returns a prompt with the following structure:
1. System prompt
2. Message history of the agent, truncated & prepended with running summary as needed
3. `cycle_instruction`
Params:
cycle_instruction: The final instruction for a thinking cycle
"""
if not cycle_instruction:
raise ValueError("No instruction given")
cycle_instruction_msg = Message("user", cycle_instruction)
cycle_instruction_tlength = count_message_tokens(
cycle_instruction_msg, self.llm.name
)
prompt = self.construct_base_prompt(reserve_tokens=cycle_instruction_tlength)
# ADD user input message ("triggering prompt")
prompt.append(cycle_instruction_msg)
return prompt
def on_before_think(self, prompt: ChatSequence, instruction: str) -> ChatSequence:
"""Called after constructing the prompt but before executing it.
Calls the `on_planning` hook of any enabled and capable plugins, adding their
output to the prompt.
Params:
instruction: The instruction for the current cycle, also used in constructing the prompt
Returns:
The prompt to execute
"""
current_tokens_used = prompt.token_length
plugin_count = len(self.config.plugins)
for i, plugin in enumerate(self.config.plugins):
if not plugin.can_handle_on_planning():
continue
plugin_response = plugin.on_planning(
self.ai_config.prompt_generator, prompt.raw()
)
if not plugin_response or plugin_response == "":
continue
message_to_add = Message("system", plugin_response)
tokens_to_add = count_message_tokens(message_to_add, self.llm.name)
if current_tokens_used + tokens_to_add > self.send_token_limit:
logger.debug(f"Plugin response too long, skipping: {plugin_response}")
logger.debug(f"Plugins remaining at stop: {plugin_count - i}")
break
prompt.insert(
-1, message_to_add
) # HACK: assumes cycle instruction to be at the end
current_tokens_used += tokens_to_add
return prompt
def on_response(
self, llm_response: ChatModelResponse, prompt: ChatSequence, instruction: str
) -> tuple[CommandName | None, CommandArgs | None, AgentThoughts]:
"""Called upon receiving a response from the chat model.
Adds the last/newest message in the prompt and the response to `history`,
and calls `self.parse_and_process_response()` to do the rest.
Params:
llm_response: The raw response from the chat model
prompt: The prompt that was executed
instruction: The instruction for the current cycle, also used in constructing the prompt
Returns:
The parsed command name and command args, if any, and the agent thoughts.
"""
# Save assistant reply to message history
self.history.append(prompt[-1])
self.history.add(
"assistant", llm_response.content, "ai_response"
) # FIXME: support function calls
try:
return self.parse_and_process_response(llm_response, prompt, instruction)
except SyntaxError as e:
logger.error(f"Response could not be parsed: {e}")
# TODO: tune this message
self.history.add(
"system",
f"Your response could not be parsed: {e}"
"\n\nRemember to only respond using the specified format above!",
)
return None, None, {}
# TODO: update memory/context
@abstractmethod
def parse_and_process_response(
self, llm_response: ChatModelResponse, prompt: ChatSequence, instruction: str
) -> tuple[CommandName | None, CommandArgs | None, AgentThoughts]:
"""Validate, parse & process the LLM's response.
Must be implemented by derivative classes: no base implementation is provided,
since the implementation depends on the role of the derivative Agent.
Params:
llm_response: The raw response from the chat model
prompt: The prompt that was executed
instruction: The instruction for the current cycle, also used in constructing the prompt
Returns:
The parsed command name and command args, if any, and the agent thoughts.
"""
pass
def add_history_upto_token_limit(
prompt: ChatSequence, history: MessageHistory, t_limit: int
) -> list[Message]:
current_prompt_length = prompt.token_length
insertion_index = len(prompt)
limit_reached = False
trimmed_messages: list[Message] = []
for cycle in reversed(list(history.per_cycle())):
messages_to_add = [msg for msg in cycle if msg is not None]
tokens_to_add = count_message_tokens(messages_to_add, prompt.model.name)
if current_prompt_length + tokens_to_add > t_limit:
limit_reached = True
if not limit_reached:
# Add the most recent message to the start of the chain,
# after the system prompts.
prompt.insert(insertion_index, *messages_to_add)
current_prompt_length += tokens_to_add
else:
trimmed_messages = messages_to_add + trimmed_messages
return trimmed_messages

147
autogpt/app/cli.py Normal file
View File

@@ -0,0 +1,147 @@
"""Main script for the autogpt package."""
from pathlib import Path
from typing import Optional
import click
@click.group(invoke_without_command=True)
@click.option("-c", "--continuous", is_flag=True, help="Enable Continuous Mode")
@click.option(
"--skip-reprompt",
"-y",
is_flag=True,
help="Skips the re-prompting messages at the beginning of the script",
)
@click.option(
"--ai-settings",
"-C",
help=(
"Specifies which ai_settings.yaml file to use, relative to the Auto-GPT"
" root directory. Will also automatically skip the re-prompt."
),
)
@click.option(
"--prompt-settings",
"-P",
help="Specifies which prompt_settings.yaml file to use.",
)
@click.option(
"-l",
"--continuous-limit",
type=int,
help="Defines the number of times to run in continuous mode",
)
@click.option("--speak", is_flag=True, help="Enable Speak Mode")
@click.option("--debug", is_flag=True, help="Enable Debug Mode")
@click.option("--gpt3only", is_flag=True, help="Enable GPT3.5 Only Mode")
@click.option("--gpt4only", is_flag=True, help="Enable GPT4 Only Mode")
@click.option(
"--use-memory",
"-m",
"memory_type",
type=str,
help="Defines which Memory backend to use",
)
@click.option(
"-b",
"--browser-name",
help="Specifies which web-browser to use when using selenium to scrape the web.",
)
@click.option(
"--allow-downloads",
is_flag=True,
help="Dangerous: Allows Auto-GPT to download files natively.",
)
@click.option(
"--skip-news",
is_flag=True,
help="Specifies whether to suppress the output of latest news on startup.",
)
@click.option(
# TODO: this is a hidden option for now, necessary for integration testing.
# We should make this public once we're ready to roll out agent specific workspaces.
"--workspace-directory",
"-w",
type=click.Path(),
hidden=True,
)
@click.option(
"--install-plugin-deps",
is_flag=True,
help="Installs external dependencies for 3rd party plugins.",
)
@click.option(
"--ai-name",
type=str,
help="AI name override",
)
@click.option(
"--ai-role",
type=str,
help="AI role override",
)
@click.option(
"--ai-goal",
type=str,
multiple=True,
help="AI goal override; may be used multiple times to pass multiple goals",
)
@click.pass_context
def main(
ctx: click.Context,
continuous: bool,
continuous_limit: int,
ai_settings: str,
prompt_settings: str,
skip_reprompt: bool,
speak: bool,
debug: bool,
gpt3only: bool,
gpt4only: bool,
memory_type: str,
browser_name: str,
allow_downloads: bool,
skip_news: bool,
workspace_directory: str,
install_plugin_deps: bool,
ai_name: Optional[str],
ai_role: Optional[str],
ai_goal: tuple[str],
) -> None:
"""
Welcome to AutoGPT an experimental open-source application showcasing the capabilities of the GPT-4 pushing the boundaries of AI.
Start an Auto-GPT assistant.
"""
# Put imports inside function to avoid importing everything when starting the CLI
from autogpt.app.main import run_auto_gpt
if ctx.invoked_subcommand is None:
run_auto_gpt(
continuous=continuous,
continuous_limit=continuous_limit,
ai_settings=ai_settings,
prompt_settings=prompt_settings,
skip_reprompt=skip_reprompt,
speak=speak,
debug=debug,
gpt3only=gpt3only,
gpt4only=gpt4only,
memory_type=memory_type,
browser_name=browser_name,
allow_downloads=allow_downloads,
skip_news=skip_news,
working_directory=Path(
__file__
).parent.parent.parent, # TODO: make this an option
workspace_directory=workspace_directory,
install_plugin_deps=install_plugin_deps,
ai_name=ai_name,
ai_role=ai_role,
ai_goals=ai_goal,
)
if __name__ == "__main__":
main()

187
autogpt/app/configurator.py Normal file
View File

@@ -0,0 +1,187 @@
"""Configurator module."""
from __future__ import annotations
from typing import Literal
import click
from colorama import Back, Fore, Style
from autogpt import utils
from autogpt.config import Config
from autogpt.config.config import GPT_3_MODEL, GPT_4_MODEL
from autogpt.llm.api_manager import ApiManager
from autogpt.logs import logger
from autogpt.memory.vector import get_supported_memory_backends
def create_config(
config: Config,
continuous: bool,
continuous_limit: int,
ai_settings_file: str,
prompt_settings_file: str,
skip_reprompt: bool,
speak: bool,
debug: bool,
gpt3only: bool,
gpt4only: bool,
memory_type: str,
browser_name: str,
allow_downloads: bool,
skip_news: bool,
) -> None:
"""Updates the config object with the given arguments.
Args:
continuous (bool): Whether to run in continuous mode
continuous_limit (int): The number of times to run in continuous mode
ai_settings_file (str): The path to the ai_settings.yaml file
prompt_settings_file (str): The path to the prompt_settings.yaml file
skip_reprompt (bool): Whether to skip the re-prompting messages at the beginning of the script
speak (bool): Whether to enable speak mode
debug (bool): Whether to enable debug mode
gpt3only (bool): Whether to enable GPT3.5 only mode
gpt4only (bool): Whether to enable GPT4 only mode
memory_type (str): The type of memory backend to use
browser_name (str): The name of the browser to use when using selenium to scrape the web
allow_downloads (bool): Whether to allow Auto-GPT to download files natively
skips_news (bool): Whether to suppress the output of latest news on startup
"""
config.debug_mode = False
config.continuous_mode = False
config.speak_mode = False
if debug:
logger.typewriter_log("Debug Mode: ", Fore.GREEN, "ENABLED")
config.debug_mode = True
if continuous:
logger.typewriter_log("Continuous Mode: ", Fore.RED, "ENABLED")
logger.typewriter_log(
"WARNING: ",
Fore.RED,
"Continuous mode is not recommended. It is potentially dangerous and may"
" cause your AI to run forever or carry out actions you would not usually"
" authorise. Use at your own risk.",
)
config.continuous_mode = True
if continuous_limit:
logger.typewriter_log(
"Continuous Limit: ", Fore.GREEN, f"{continuous_limit}"
)
config.continuous_limit = continuous_limit
# Check if continuous limit is used without continuous mode
if continuous_limit and not continuous:
raise click.UsageError("--continuous-limit can only be used with --continuous")
if speak:
logger.typewriter_log("Speak Mode: ", Fore.GREEN, "ENABLED")
config.speak_mode = True
# Set the default LLM models
if gpt3only:
logger.typewriter_log("GPT3.5 Only Mode: ", Fore.GREEN, "ENABLED")
# --gpt3only should always use gpt-3.5-turbo, despite user's FAST_LLM config
config.fast_llm = GPT_3_MODEL
config.smart_llm = GPT_3_MODEL
elif (
gpt4only
and check_model(GPT_4_MODEL, model_type="smart_llm", config=config)
== GPT_4_MODEL
):
logger.typewriter_log("GPT4 Only Mode: ", Fore.GREEN, "ENABLED")
# --gpt4only should always use gpt-4, despite user's SMART_LLM config
config.fast_llm = GPT_4_MODEL
config.smart_llm = GPT_4_MODEL
else:
config.fast_llm = check_model(config.fast_llm, "fast_llm", config=config)
config.smart_llm = check_model(config.smart_llm, "smart_llm", config=config)
if memory_type:
supported_memory = get_supported_memory_backends()
chosen = memory_type
if chosen not in supported_memory:
logger.typewriter_log(
"ONLY THE FOLLOWING MEMORY BACKENDS ARE SUPPORTED: ",
Fore.RED,
f"{supported_memory}",
)
logger.typewriter_log("Defaulting to: ", Fore.YELLOW, config.memory_backend)
else:
config.memory_backend = chosen
if skip_reprompt:
logger.typewriter_log("Skip Re-prompt: ", Fore.GREEN, "ENABLED")
config.skip_reprompt = True
if ai_settings_file:
file = ai_settings_file
# Validate file
(validated, message) = utils.validate_yaml_file(file)
if not validated:
logger.typewriter_log("FAILED FILE VALIDATION", Fore.RED, message)
logger.double_check()
exit(1)
logger.typewriter_log("Using AI Settings File:", Fore.GREEN, file)
config.ai_settings_file = file
config.skip_reprompt = True
if prompt_settings_file:
file = prompt_settings_file
# Validate file
(validated, message) = utils.validate_yaml_file(file)
if not validated:
logger.typewriter_log("FAILED FILE VALIDATION", Fore.RED, message)
logger.double_check()
exit(1)
logger.typewriter_log("Using Prompt Settings File:", Fore.GREEN, file)
config.prompt_settings_file = file
if browser_name:
config.selenium_web_browser = browser_name
if allow_downloads:
logger.typewriter_log("Native Downloading:", Fore.GREEN, "ENABLED")
logger.typewriter_log(
"WARNING: ",
Fore.YELLOW,
f"{Back.LIGHTYELLOW_EX}Auto-GPT will now be able to download and save files to your machine.{Back.RESET} "
+ "It is recommended that you monitor any files it downloads carefully.",
)
logger.typewriter_log(
"WARNING: ",
Fore.YELLOW,
f"{Back.RED + Style.BRIGHT}ALWAYS REMEMBER TO NEVER OPEN FILES YOU AREN'T SURE OF!{Style.RESET_ALL}",
)
config.allow_downloads = True
if skip_news:
config.skip_news = True
def check_model(
model_name: str,
model_type: Literal["smart_llm", "fast_llm"],
config: Config,
) -> str:
"""Check if model is available for use. If not, return gpt-3.5-turbo."""
openai_credentials = config.get_openai_credentials(model_name)
api_manager = ApiManager()
models = api_manager.get_models(**openai_credentials)
if any(model_name in m["id"] for m in models):
return model_name
logger.typewriter_log(
"WARNING: ",
Fore.YELLOW,
f"You do not have access to {model_name}. Setting {model_type} to "
f"gpt-3.5-turbo.",
)
return "gpt-3.5-turbo"

597
autogpt/app/main.py Normal file
View File

@@ -0,0 +1,597 @@
"""The application entry point. Can be invoked by a CLI or any other front end application."""
import enum
import logging
import math
import signal
import sys
from pathlib import Path
from types import FrameType
from typing import Optional
from colorama import Fore, Style
from autogpt.agents import Agent, AgentThoughts, CommandArgs, CommandName
from autogpt.app.configurator import create_config
from autogpt.app.setup import prompt_user
from autogpt.commands import COMMAND_CATEGORIES
from autogpt.config import AIConfig, Config, ConfigBuilder, check_openai_api_key
from autogpt.llm.api_manager import ApiManager
from autogpt.logs import logger
from autogpt.memory.vector import get_memory
from autogpt.models.command_registry import CommandRegistry
from autogpt.plugins import scan_plugins
from autogpt.prompts.prompt import DEFAULT_TRIGGERING_PROMPT
from autogpt.speech import say_text
from autogpt.spinner import Spinner
from autogpt.utils import (
clean_input,
get_current_git_branch,
get_latest_bulletin,
get_legal_warning,
markdown_to_ansi_style,
)
from autogpt.workspace import Workspace
from scripts.install_plugin_deps import install_plugin_dependencies
def run_auto_gpt(
continuous: bool,
continuous_limit: int,
ai_settings: str,
prompt_settings: str,
skip_reprompt: bool,
speak: bool,
debug: bool,
gpt3only: bool,
gpt4only: bool,
memory_type: str,
browser_name: str,
allow_downloads: bool,
skip_news: bool,
working_directory: Path,
workspace_directory: str | Path,
install_plugin_deps: bool,
ai_name: Optional[str] = None,
ai_role: Optional[str] = None,
ai_goals: tuple[str] = tuple(),
):
# Configure logging before we do anything else.
logger.set_level(logging.DEBUG if debug else logging.INFO)
config = ConfigBuilder.build_config_from_env(workdir=working_directory)
# HACK: This is a hack to allow the config into the logger without having to pass it around everywhere
# or import it directly.
logger.config = config
# TODO: fill in llm values here
check_openai_api_key(config)
create_config(
config,
continuous,
continuous_limit,
ai_settings,
prompt_settings,
skip_reprompt,
speak,
debug,
gpt3only,
gpt4only,
memory_type,
browser_name,
allow_downloads,
skip_news,
)
if config.continuous_mode:
for line in get_legal_warning().split("\n"):
logger.warn(markdown_to_ansi_style(line), "LEGAL:", Fore.RED)
if not config.skip_news:
motd, is_new_motd = get_latest_bulletin()
if motd:
motd = markdown_to_ansi_style(motd)
for motd_line in motd.split("\n"):
logger.info(motd_line, "NEWS:", Fore.GREEN)
if is_new_motd and not config.chat_messages_enabled:
input(
Fore.MAGENTA
+ Style.BRIGHT
+ "NEWS: Bulletin was updated! Press Enter to continue..."
+ Style.RESET_ALL
)
git_branch = get_current_git_branch()
if git_branch and git_branch != "stable":
logger.typewriter_log(
"WARNING: ",
Fore.RED,
f"You are running on `{git_branch}` branch "
"- this is not a supported branch.",
)
if sys.version_info < (3, 10):
logger.typewriter_log(
"WARNING: ",
Fore.RED,
"You are running on an older version of Python. "
"Some people have observed problems with certain "
"parts of Auto-GPT with this version. "
"Please consider upgrading to Python 3.10 or higher.",
)
if install_plugin_deps:
install_plugin_dependencies()
# TODO: have this directory live outside the repository (e.g. in a user's
# home directory) and have it come in as a command line argument or part of
# the env file.
Workspace.set_workspace_directory(config, workspace_directory)
# HACK: doing this here to collect some globals that depend on the workspace.
Workspace.set_file_logger_path(config, config.workspace_path)
config.plugins = scan_plugins(config, config.debug_mode)
# Create a CommandRegistry instance and scan default folder
command_registry = CommandRegistry()
logger.debug(
f"The following command categories are disabled: {config.disabled_command_categories}"
)
enabled_command_categories = [
x for x in COMMAND_CATEGORIES if x not in config.disabled_command_categories
]
logger.debug(
f"The following command categories are enabled: {enabled_command_categories}"
)
for command_category in enabled_command_categories:
command_registry.import_commands(command_category)
# Unregister commands that are incompatible with the current config
incompatible_commands = []
for command in command_registry.commands.values():
if callable(command.enabled) and not command.enabled(config):
command.enabled = False
incompatible_commands.append(command)
for command in incompatible_commands:
command_registry.unregister(command)
logger.debug(
f"Unregistering incompatible command: {command.name}, "
f"reason - {command.disabled_reason or 'Disabled by current config.'}"
)
ai_config = construct_main_ai_config(
config,
name=ai_name,
role=ai_role,
goals=ai_goals,
)
ai_config.command_registry = command_registry
# print(prompt)
# add chat plugins capable of report to logger
if config.chat_messages_enabled:
for plugin in config.plugins:
if hasattr(plugin, "can_handle_report") and plugin.can_handle_report():
logger.info(f"Loaded plugin into logger: {plugin.__class__.__name__}")
logger.chat_plugins.append(plugin)
# Initialize memory and make sure it is empty.
# this is particularly important for indexing and referencing pinecone memory
memory = get_memory(config)
memory.clear()
logger.typewriter_log(
"Using memory of type:", Fore.GREEN, f"{memory.__class__.__name__}"
)
logger.typewriter_log("Using Browser:", Fore.GREEN, config.selenium_web_browser)
agent = Agent(
memory=memory,
command_registry=command_registry,
triggering_prompt=DEFAULT_TRIGGERING_PROMPT,
ai_config=ai_config,
config=config,
)
run_interaction_loop(agent)
def _get_cycle_budget(continuous_mode: bool, continuous_limit: int) -> int | None:
# Translate from the continuous_mode/continuous_limit config
# to a cycle_budget (maximum number of cycles to run without checking in with the
# user) and a count of cycles_remaining before we check in..
if continuous_mode:
cycle_budget = continuous_limit if continuous_limit else math.inf
else:
cycle_budget = 1
return cycle_budget
class UserFeedback(str, enum.Enum):
"""Enum for user feedback."""
AUTHORIZE = "GENERATE NEXT COMMAND JSON"
EXIT = "EXIT"
TEXT = "TEXT"
def run_interaction_loop(
agent: Agent,
) -> None:
"""Run the main interaction loop for the agent.
Args:
agent: The agent to run the interaction loop for.
Returns:
None
"""
# These contain both application config and agent config, so grab them here.
config = agent.config
ai_config = agent.ai_config
logger.debug(f"{ai_config.ai_name} System Prompt: {agent.system_prompt}")
cycle_budget = cycles_remaining = _get_cycle_budget(
config.continuous_mode, config.continuous_limit
)
spinner = Spinner("Thinking...", plain_output=config.plain_output)
def graceful_agent_interrupt(signum: int, frame: Optional[FrameType]) -> None:
nonlocal cycle_budget, cycles_remaining, spinner
if cycles_remaining in [0, 1, math.inf]:
logger.typewriter_log(
"Interrupt signal received. Stopping continuous command execution "
"immediately.",
Fore.RED,
)
sys.exit()
else:
restart_spinner = spinner.running
if spinner.running:
spinner.stop()
logger.typewriter_log(
"Interrupt signal received. Stopping continuous command execution.",
Fore.RED,
)
cycles_remaining = 1
if restart_spinner:
spinner.start()
# Set up an interrupt signal for the agent.
signal.signal(signal.SIGINT, graceful_agent_interrupt)
#########################
# Application Main Loop #
#########################
while cycles_remaining > 0:
logger.debug(f"Cycle budget: {cycle_budget}; remaining: {cycles_remaining}")
########
# Plan #
########
# Have the agent determine the next action to take.
with spinner:
command_name, command_args, assistant_reply_dict = agent.think()
###############
# Update User #
###############
# Print the assistant's thoughts and the next command to the user.
update_user(config, ai_config, command_name, command_args, assistant_reply_dict)
##################
# Get user input #
##################
if cycles_remaining == 1: # Last cycle
user_feedback, user_input, new_cycles_remaining = get_user_feedback(
config,
ai_config,
)
if user_feedback == UserFeedback.AUTHORIZE:
if new_cycles_remaining is not None:
# Case 1: User is altering the cycle budget.
if cycle_budget > 1:
cycle_budget = new_cycles_remaining + 1
# Case 2: User is running iteratively and
# has initiated a one-time continuous cycle
cycles_remaining = new_cycles_remaining + 1
else:
# Case 1: Continuous iteration was interrupted -> resume
if cycle_budget > 1:
logger.typewriter_log(
"RESUMING CONTINUOUS EXECUTION: ",
Fore.MAGENTA,
f"The cycle budget is {cycle_budget}.",
)
# Case 2: The agent used up its cycle budget -> reset
cycles_remaining = cycle_budget + 1
logger.typewriter_log(
"-=-=-=-=-=-=-= COMMAND AUTHORISED BY USER -=-=-=-=-=-=-=",
Fore.MAGENTA,
"",
)
elif user_feedback == UserFeedback.EXIT:
logger.typewriter_log("Exiting...", Fore.YELLOW)
exit()
else: # user_feedback == UserFeedback.TEXT
command_name = "human_feedback"
else:
user_input = None
# First log new-line so user can differentiate sections better in console
logger.typewriter_log("\n")
if cycles_remaining != math.inf:
# Print authorized commands left value
logger.typewriter_log(
"AUTHORISED COMMANDS LEFT: ", Fore.CYAN, f"{cycles_remaining}"
)
###################
# Execute Command #
###################
# Decrement the cycle counter first to reduce the likelihood of a SIGINT
# happening during command execution, setting the cycles remaining to 1,
# and then having the decrement set it to 0, exiting the application.
if command_name != "human_feedback":
cycles_remaining -= 1
result = agent.execute(command_name, command_args, user_input)
if result is not None:
logger.typewriter_log("SYSTEM: ", Fore.YELLOW, result)
else:
logger.typewriter_log("SYSTEM: ", Fore.YELLOW, "Unable to execute command")
def update_user(
config: Config,
ai_config: AIConfig,
command_name: CommandName | None,
command_args: CommandArgs | None,
assistant_reply_dict: AgentThoughts,
) -> None:
"""Prints the assistant's thoughts and the next command to the user.
Args:
config: The program's configuration.
ai_config: The AI's configuration.
command_name: The name of the command to execute.
command_args: The arguments for the command.
assistant_reply_dict: The assistant's reply.
"""
print_assistant_thoughts(ai_config.ai_name, assistant_reply_dict, config)
if command_name is not None:
if config.speak_mode:
say_text(f"I want to execute {command_name}", config)
# First log new-line so user can differentiate sections better in console
logger.typewriter_log("\n")
logger.typewriter_log(
"NEXT ACTION: ",
Fore.CYAN,
f"COMMAND = {Fore.CYAN}{remove_ansi_escape(command_name)}{Style.RESET_ALL} "
f"ARGUMENTS = {Fore.CYAN}{command_args}{Style.RESET_ALL}",
)
elif command_name.lower().startswith("error"):
logger.typewriter_log(
"ERROR: ",
Fore.RED,
f"The Agent failed to select an action. " f"Error message: {command_name}",
)
else:
logger.typewriter_log(
"NO ACTION SELECTED: ",
Fore.RED,
f"The Agent failed to select an action.",
)
def get_user_feedback(
config: Config,
ai_config: AIConfig,
) -> tuple[UserFeedback, str, int | None]:
"""Gets the user's feedback on the assistant's reply.
Args:
config: The program's configuration.
ai_config: The AI's configuration.
Returns:
A tuple of the user's feedback, the user's input, and the number of
cycles remaining if the user has initiated a continuous cycle.
"""
# ### GET USER AUTHORIZATION TO EXECUTE COMMAND ###
# Get key press: Prompt the user to press enter to continue or escape
# to exit
logger.info(
f"Enter '{config.authorise_key}' to authorise command, "
f"'{config.authorise_key} -N' to run N continuous commands, "
f"'{config.exit_key}' to exit program, or enter feedback for "
f"{ai_config.ai_name}..."
)
user_feedback = None
user_input = ""
new_cycles_remaining = None
while user_feedback is None:
# Get input from user
if config.chat_messages_enabled:
console_input = clean_input(config, "Waiting for your response...")
else:
console_input = clean_input(
config, Fore.MAGENTA + "Input:" + Style.RESET_ALL
)
# Parse user input
if console_input.lower().strip() == config.authorise_key:
user_feedback = UserFeedback.AUTHORIZE
elif console_input.lower().strip() == "":
logger.warn("Invalid input format.")
elif console_input.lower().startswith(f"{config.authorise_key} -"):
try:
user_feedback = UserFeedback.AUTHORIZE
new_cycles_remaining = abs(int(console_input.split(" ")[1]))
except ValueError:
logger.warn(
f"Invalid input format. "
f"Please enter '{config.authorise_key} -N'"
" where N is the number of continuous tasks."
)
elif console_input.lower() in [config.exit_key, "exit"]:
user_feedback = UserFeedback.EXIT
else:
user_feedback = UserFeedback.TEXT
user_input = console_input
return user_feedback, user_input, new_cycles_remaining
def construct_main_ai_config(
config: Config,
name: Optional[str] = None,
role: Optional[str] = None,
goals: tuple[str] = tuple(),
) -> AIConfig:
"""Construct the prompt for the AI to respond to
Returns:
str: The prompt string
"""
ai_config = AIConfig.load(config.workdir / config.ai_settings_file)
# Apply overrides
if name:
ai_config.ai_name = name
if role:
ai_config.ai_role = role
if goals:
ai_config.ai_goals = list(goals)
if (
all([name, role, goals])
or config.skip_reprompt
and all([ai_config.ai_name, ai_config.ai_role, ai_config.ai_goals])
):
logger.typewriter_log("Name :", Fore.GREEN, ai_config.ai_name)
logger.typewriter_log("Role :", Fore.GREEN, ai_config.ai_role)
logger.typewriter_log("Goals:", Fore.GREEN, f"{ai_config.ai_goals}")
logger.typewriter_log(
"API Budget:",
Fore.GREEN,
"infinite" if ai_config.api_budget <= 0 else f"${ai_config.api_budget}",
)
elif all([ai_config.ai_name, ai_config.ai_role, ai_config.ai_goals]):
logger.typewriter_log(
"Welcome back! ",
Fore.GREEN,
f"Would you like me to return to being {ai_config.ai_name}?",
speak_text=True,
)
should_continue = clean_input(
config,
f"""Continue with the last settings?
Name: {ai_config.ai_name}
Role: {ai_config.ai_role}
Goals: {ai_config.ai_goals}
API Budget: {"infinite" if ai_config.api_budget <= 0 else f"${ai_config.api_budget}"}
Continue ({config.authorise_key}/{config.exit_key}): """,
)
if should_continue.lower() == config.exit_key:
ai_config = AIConfig()
if any([not ai_config.ai_name, not ai_config.ai_role, not ai_config.ai_goals]):
ai_config = prompt_user(config)
ai_config.save(config.workdir / config.ai_settings_file)
if config.restrict_to_workspace:
logger.typewriter_log(
"NOTE:All files/directories created by this agent can be found inside its workspace at:",
Fore.YELLOW,
f"{config.workspace_path}",
)
# set the total api budget
api_manager = ApiManager()
api_manager.set_total_budget(ai_config.api_budget)
# Agent Created, print message
logger.typewriter_log(
ai_config.ai_name,
Fore.LIGHTBLUE_EX,
"has been created with the following details:",
speak_text=True,
)
# Print the ai_config details
# Name
logger.typewriter_log("Name:", Fore.GREEN, ai_config.ai_name, speak_text=False)
# Role
logger.typewriter_log("Role:", Fore.GREEN, ai_config.ai_role, speak_text=False)
# Goals
logger.typewriter_log("Goals:", Fore.GREEN, "", speak_text=False)
for goal in ai_config.ai_goals:
logger.typewriter_log("-", Fore.GREEN, goal, speak_text=False)
return ai_config
def print_assistant_thoughts(
ai_name: str,
assistant_reply_json_valid: dict,
config: Config,
) -> None:
from autogpt.speech import say_text
assistant_thoughts_reasoning = None
assistant_thoughts_plan = None
assistant_thoughts_speak = None
assistant_thoughts_criticism = None
assistant_thoughts = assistant_reply_json_valid.get("thoughts", {})
assistant_thoughts_text = remove_ansi_escape(assistant_thoughts.get("text", ""))
if assistant_thoughts:
assistant_thoughts_reasoning = remove_ansi_escape(
assistant_thoughts.get("reasoning", "")
)
assistant_thoughts_plan = remove_ansi_escape(assistant_thoughts.get("plan", ""))
assistant_thoughts_criticism = remove_ansi_escape(
assistant_thoughts.get("criticism", "")
)
assistant_thoughts_speak = remove_ansi_escape(
assistant_thoughts.get("speak", "")
)
logger.typewriter_log(
f"{ai_name.upper()} THOUGHTS:", Fore.YELLOW, assistant_thoughts_text
)
logger.typewriter_log("REASONING:", Fore.YELLOW, str(assistant_thoughts_reasoning))
if assistant_thoughts_plan:
logger.typewriter_log("PLAN:", Fore.YELLOW, "")
# If it's a list, join it into a string
if isinstance(assistant_thoughts_plan, list):
assistant_thoughts_plan = "\n".join(assistant_thoughts_plan)
elif isinstance(assistant_thoughts_plan, dict):
assistant_thoughts_plan = str(assistant_thoughts_plan)
# Split the input_string using the newline character and dashes
lines = assistant_thoughts_plan.split("\n")
for line in lines:
line = line.lstrip("- ")
logger.typewriter_log("- ", Fore.GREEN, line.strip())
logger.typewriter_log("CRITICISM:", Fore.YELLOW, f"{assistant_thoughts_criticism}")
# Speak the assistant's thoughts
if assistant_thoughts_speak:
if config.speak_mode:
say_text(assistant_thoughts_speak, config)
else:
logger.typewriter_log("SPEAK:", Fore.YELLOW, f"{assistant_thoughts_speak}")
def remove_ansi_escape(s: str) -> str:
return s.replace("\x1B", "")

238
autogpt/app/setup.py Normal file
View File

@@ -0,0 +1,238 @@
"""Set up the AI and its goals"""
import re
from typing import Optional
from colorama import Fore, Style
from jinja2 import Template
from autogpt import utils
from autogpt.config import Config
from autogpt.config.ai_config import AIConfig
from autogpt.llm.base import ChatSequence, Message
from autogpt.llm.utils import create_chat_completion
from autogpt.logs import logger
from autogpt.prompts.default_prompts import (
DEFAULT_SYSTEM_PROMPT_AICONFIG_AUTOMATIC,
DEFAULT_TASK_PROMPT_AICONFIG_AUTOMATIC,
DEFAULT_USER_DESIRE_PROMPT,
)
def prompt_user(
config: Config, ai_config_template: Optional[AIConfig] = None
) -> AIConfig:
"""Prompt the user for input
Params:
config (Config): The Config object
ai_config_template (AIConfig): The AIConfig object to use as a template
Returns:
AIConfig: The AIConfig object tailored to the user's input
"""
# Construct the prompt
logger.typewriter_log(
"Welcome to Auto-GPT! ",
Fore.GREEN,
"run with '--help' for more information.",
speak_text=True,
)
ai_config_template_provided = ai_config_template is not None and any(
[
ai_config_template.ai_goals,
ai_config_template.ai_name,
ai_config_template.ai_role,
]
)
user_desire = ""
if not ai_config_template_provided:
# Get user desire if command line overrides have not been passed in
logger.typewriter_log(
"Create an AI-Assistant:",
Fore.GREEN,
"input '--manual' to enter manual mode.",
speak_text=True,
)
user_desire = utils.clean_input(
config, f"{Fore.LIGHTBLUE_EX}I want Auto-GPT to{Style.RESET_ALL}: "
)
if user_desire.strip() == "":
user_desire = DEFAULT_USER_DESIRE_PROMPT # Default prompt
# If user desire contains "--manual" or we have overridden any of the AI configuration
if "--manual" in user_desire or ai_config_template_provided:
logger.typewriter_log(
"Manual Mode Selected",
Fore.GREEN,
speak_text=True,
)
return generate_aiconfig_manual(config, ai_config_template)
else:
try:
return generate_aiconfig_automatic(user_desire, config)
except Exception as e:
logger.typewriter_log(
"Unable to automatically generate AI Config based on user desire.",
Fore.RED,
"Falling back to manual mode.",
speak_text=True,
)
return generate_aiconfig_manual(config)
def generate_aiconfig_manual(
config: Config, ai_config_template: Optional[AIConfig] = None
) -> AIConfig:
"""
Interactively create an AI configuration by prompting the user to provide the name, role, and goals of the AI.
This function guides the user through a series of prompts to collect the necessary information to create
an AIConfig object. The user will be asked to provide a name and role for the AI, as well as up to five
goals. If the user does not provide a value for any of the fields, default values will be used.
Params:
config (Config): The Config object
ai_config_template (AIConfig): The AIConfig object to use as a template
Returns:
AIConfig: An AIConfig object containing the user-defined or default AI name, role, and goals.
"""
# Manual Setup Intro
logger.typewriter_log(
"Create an AI-Assistant:",
Fore.GREEN,
"Enter the name of your AI and its role below. Entering nothing will load"
" defaults.",
speak_text=True,
)
if ai_config_template and ai_config_template.ai_name:
ai_name = ai_config_template.ai_name
else:
ai_name = ""
# Get AI Name from User
logger.typewriter_log(
"Name your AI: ", Fore.GREEN, "For example, 'Entrepreneur-GPT'"
)
ai_name = utils.clean_input(config, "AI Name: ")
if ai_name == "":
ai_name = "Entrepreneur-GPT"
logger.typewriter_log(
f"{ai_name} here!", Fore.LIGHTBLUE_EX, "I am at your service.", speak_text=True
)
if ai_config_template and ai_config_template.ai_role:
ai_role = ai_config_template.ai_role
else:
# Get AI Role from User
logger.typewriter_log(
"Describe your AI's role: ",
Fore.GREEN,
"For example, 'an AI designed to autonomously develop and run businesses with"
" the sole goal of increasing your net worth.'",
)
ai_role = utils.clean_input(config, f"{ai_name} is: ")
if ai_role == "":
ai_role = "an AI designed to autonomously develop and run businesses with the"
" sole goal of increasing your net worth."
if ai_config_template and ai_config_template.ai_goals:
ai_goals = ai_config_template.ai_goals
else:
# Enter up to 5 goals for the AI
logger.typewriter_log(
"Enter up to 5 goals for your AI: ",
Fore.GREEN,
"For example: \nIncrease net worth, Grow Twitter Account, Develop and manage"
" multiple businesses autonomously'",
)
logger.info("Enter nothing to load defaults, enter nothing when finished.")
ai_goals = []
for i in range(5):
ai_goal = utils.clean_input(
config, f"{Fore.LIGHTBLUE_EX}Goal{Style.RESET_ALL} {i+1}: "
)
if ai_goal == "":
break
ai_goals.append(ai_goal)
if not ai_goals:
ai_goals = [
"Increase net worth",
"Grow Twitter Account",
"Develop and manage multiple businesses autonomously",
]
# Get API Budget from User
logger.typewriter_log(
"Enter your budget for API calls: ",
Fore.GREEN,
"For example: $1.50",
)
logger.info("Enter nothing to let the AI run without monetary limit")
api_budget_input = utils.clean_input(
config, f"{Fore.LIGHTBLUE_EX}Budget{Style.RESET_ALL}: $"
)
if api_budget_input == "":
api_budget = 0.0
else:
try:
api_budget = float(api_budget_input.replace("$", ""))
except ValueError:
logger.typewriter_log(
"Invalid budget input. Setting budget to unlimited.", Fore.RED
)
api_budget = 0.0
return AIConfig(ai_name, ai_role, ai_goals, api_budget)
def generate_aiconfig_automatic(user_prompt: str, config: Config) -> AIConfig:
"""Generates an AIConfig object from the given string.
Returns:
AIConfig: The AIConfig object tailored to the user's input
"""
system_prompt = DEFAULT_SYSTEM_PROMPT_AICONFIG_AUTOMATIC
prompt_ai_config_automatic = Template(
DEFAULT_TASK_PROMPT_AICONFIG_AUTOMATIC
).render(user_prompt=user_prompt)
# Call LLM with the string as user input
output = create_chat_completion(
ChatSequence.for_model(
config.fast_llm,
[
Message("system", system_prompt),
Message("user", prompt_ai_config_automatic),
],
),
config,
).content
# Debug LLM Output
logger.debug(f"AI Config Generator Raw Output: {output}")
# Parse the output
ai_name = re.search(r"Name(?:\s*):(?:\s*)(.*)", output, re.IGNORECASE).group(1)
ai_role = (
re.search(
r"Description(?:\s*):(?:\s*)(.*?)(?:(?:\n)|Goals)",
output,
re.IGNORECASE | re.DOTALL,
)
.group(1)
.strip()
)
ai_goals = re.findall(r"(?<=\n)-\s*(.*)", output)
api_budget = 0.0 # TODO: parse api budget using a regular expression
return AIConfig(ai_name, ai_role, ai_goals, api_budget)

View File

@@ -0,0 +1,57 @@
import functools
from typing import Any, Callable, Optional, TypedDict
from autogpt.config import Config
from autogpt.models.command import Command, CommandParameter
# Unique identifier for auto-gpt commands
AUTO_GPT_COMMAND_IDENTIFIER = "auto_gpt_command"
class CommandParameterSpec(TypedDict):
type: str
description: str
required: bool
def command(
name: str,
description: str,
parameters: dict[str, CommandParameterSpec],
enabled: bool | Callable[[Config], bool] = True,
disabled_reason: Optional[str] = None,
aliases: list[str] = [],
) -> Callable[..., Any]:
"""The command decorator is used to create Command objects from ordinary functions."""
def decorator(func: Callable[..., Any]) -> Command:
typed_parameters = [
CommandParameter(
name=param_name,
description=parameter.get("description"),
type=parameter.get("type", "string"),
required=parameter.get("required", False),
)
for param_name, parameter in parameters.items()
]
cmd = Command(
name=name,
description=description,
method=func,
parameters=typed_parameters,
enabled=enabled,
disabled_reason=disabled_reason,
aliases=aliases,
)
@functools.wraps(func)
def wrapper(*args, **kwargs) -> Any:
return func(*args, **kwargs)
wrapper.command = cmd
setattr(wrapper, AUTO_GPT_COMMAND_IDENTIFIER, True)
return wrapper
return decorator

View File

@@ -0,0 +1,7 @@
COMMAND_CATEGORIES = [
"autogpt.commands.execute_code",
"autogpt.commands.file_operations",
"autogpt.commands.web_search",
"autogpt.commands.web_selenium",
"autogpt.commands.task_statuses",
]

View File

@@ -0,0 +1,64 @@
import functools
from pathlib import Path
from typing import Callable
from autogpt.agents.agent import Agent
from autogpt.logs import logger
def sanitize_path_arg(arg_name: str):
def decorator(func: Callable):
# Get position of path parameter, in case it is passed as a positional argument
try:
arg_index = list(func.__annotations__.keys()).index(arg_name)
except ValueError:
raise TypeError(
f"Sanitized parameter '{arg_name}' absent or not annotated on function '{func.__name__}'"
)
# Get position of agent parameter, in case it is passed as a positional argument
try:
agent_arg_index = list(func.__annotations__.keys()).index("agent")
except ValueError:
raise TypeError(
f"Parameter 'agent' absent or not annotated on function '{func.__name__}'"
)
@functools.wraps(func)
def wrapper(*args, **kwargs):
logger.debug(f"Sanitizing arg '{arg_name}' on function '{func.__name__}'")
logger.debug(f"Function annotations: {func.__annotations__}")
# Get Agent from the called function's arguments
agent = kwargs.get(
"agent", len(args) > agent_arg_index and args[agent_arg_index]
)
logger.debug(f"Args: {args}")
logger.debug(f"KWArgs: {kwargs}")
logger.debug(f"Agent argument lifted from function call: {agent}")
if not isinstance(agent, Agent):
raise RuntimeError("Could not get Agent from decorated command's args")
# Sanitize the specified path argument, if one is given
given_path: str | Path | None = kwargs.get(
arg_name, len(args) > arg_index and args[arg_index] or None
)
if given_path:
if given_path in {"", "/"}:
sanitized_path = str(agent.workspace.root)
else:
sanitized_path = str(agent.workspace.get_path(given_path))
if arg_name in kwargs:
kwargs[arg_name] = sanitized_path
else:
# args is an immutable tuple; must be converted to a list to update
arg_list = list(args)
arg_list[arg_index] = sanitized_path
args = tuple(arg_list)
return func(*args, **kwargs)
return wrapper
return decorator

View File

@@ -0,0 +1,306 @@
"""Execute code in a Docker container"""
import os
import subprocess
from pathlib import Path
import docker
from docker.errors import DockerException, ImageNotFound
from docker.models.containers import Container as DockerContainer
from autogpt.agents.agent import Agent
from autogpt.command_decorator import command
from autogpt.config import Config
from autogpt.logs import logger
from .decorators import sanitize_path_arg
ALLOWLIST_CONTROL = "allowlist"
DENYLIST_CONTROL = "denylist"
@command(
"execute_python_code",
"Creates a Python file and executes it",
{
"code": {
"type": "string",
"description": "The Python code to run",
"required": True,
},
"name": {
"type": "string",
"description": "A name to be given to the python file",
"required": True,
},
},
)
def execute_python_code(code: str, name: str, agent: Agent) -> str:
"""Create and execute a Python file in a Docker container and return the STDOUT of the
executed code. If there is any data that needs to be captured use a print statement
Args:
code (str): The Python code to run
name (str): A name to be given to the Python file
Returns:
str: The STDOUT captured from the code when it ran
"""
ai_name = agent.ai_config.ai_name
code_dir = agent.workspace.get_path(Path(ai_name, "executed_code"))
os.makedirs(code_dir, exist_ok=True)
if not name.endswith(".py"):
name = name + ".py"
# The `name` arg is not covered by @sanitize_path_arg,
# so sanitization must be done here to prevent path traversal.
file_path = agent.workspace.get_path(code_dir / name)
if not file_path.is_relative_to(code_dir):
return "Error: 'name' argument resulted in path traversal, operation aborted"
try:
with open(file_path, "w+", encoding="utf-8") as f:
f.write(code)
return execute_python_file(str(file_path), agent)
except Exception as e:
return f"Error: {str(e)}"
@command(
"execute_python_file",
"Executes an existing Python file",
{
"filename": {
"type": "string",
"description": "The name of te file to execute",
"required": True,
},
},
)
@sanitize_path_arg("filename")
def execute_python_file(filename: str, agent: Agent) -> str:
"""Execute a Python file in a Docker container and return the output
Args:
filename (str): The name of the file to execute
Returns:
str: The output of the file
"""
logger.info(
f"Executing python file '{filename}' in working directory '{agent.config.workspace_path}'"
)
if not filename.endswith(".py"):
return "Error: Invalid file type. Only .py files are allowed."
file_path = Path(filename)
if not file_path.is_file():
# Mimic the response that you get from the command line so that it's easier to identify
return (
f"python: can't open file '{filename}': [Errno 2] No such file or directory"
)
if we_are_running_in_a_docker_container():
logger.debug(
f"Auto-GPT is running in a Docker container; executing {file_path} directly..."
)
result = subprocess.run(
["python", str(file_path)],
capture_output=True,
encoding="utf8",
cwd=agent.config.workspace_path,
)
if result.returncode == 0:
return result.stdout
else:
return f"Error: {result.stderr}"
logger.debug("Auto-GPT is not running in a Docker container")
try:
client = docker.from_env()
# You can replace this with the desired Python image/version
# You can find available Python images on Docker Hub:
# https://hub.docker.com/_/python
image_name = "python:3-alpine"
try:
client.images.get(image_name)
logger.debug(f"Image '{image_name}' found locally")
except ImageNotFound:
logger.info(
f"Image '{image_name}' not found locally, pulling from Docker Hub..."
)
# Use the low-level API to stream the pull response
low_level_client = docker.APIClient()
for line in low_level_client.pull(image_name, stream=True, decode=True):
# Print the status and progress, if available
status = line.get("status")
progress = line.get("progress")
if status and progress:
logger.info(f"{status}: {progress}")
elif status:
logger.info(status)
logger.debug(f"Running {file_path} in a {image_name} container...")
container: DockerContainer = client.containers.run(
image_name,
[
"python",
file_path.relative_to(agent.workspace.root).as_posix(),
],
volumes={
str(agent.config.workspace_path): {
"bind": "/workspace",
"mode": "rw",
}
},
working_dir="/workspace",
stderr=True,
stdout=True,
detach=True,
) # type: ignore
container.wait()
logs = container.logs().decode("utf-8")
container.remove()
# print(f"Execution complete. Output: {output}")
# print(f"Logs: {logs}")
return logs
except DockerException as e:
logger.warn(
"Could not run the script in a container. If you haven't already, please install Docker https://docs.docker.com/get-docker/"
)
return f"Error: {str(e)}"
except Exception as e:
return f"Error: {str(e)}"
def validate_command(command: str, config: Config) -> bool:
"""Validate a command to ensure it is allowed
Args:
command (str): The command to validate
config (Config): The config to use to validate the command
Returns:
bool: True if the command is allowed, False otherwise
"""
if not command:
return False
command_name = command.split()[0]
if config.shell_command_control == ALLOWLIST_CONTROL:
return command_name in config.shell_allowlist
else:
return command_name not in config.shell_denylist
@command(
"execute_shell",
"Executes a Shell Command, non-interactive commands only",
{
"command_line": {
"type": "string",
"description": "The command line to execute",
"required": True,
}
},
enabled=lambda config: config.execute_local_commands,
disabled_reason="You are not allowed to run local shell commands. To execute"
" shell commands, EXECUTE_LOCAL_COMMANDS must be set to 'True' "
"in your config file: .env - do not attempt to bypass the restriction.",
)
def execute_shell(command_line: str, agent: Agent) -> str:
"""Execute a shell command and return the output
Args:
command_line (str): The command line to execute
Returns:
str: The output of the command
"""
if not validate_command(command_line, agent.config):
logger.info(f"Command '{command_line}' not allowed")
return "Error: This Shell Command is not allowed."
current_dir = Path.cwd()
# Change dir into workspace if necessary
if not current_dir.is_relative_to(agent.config.workspace_path):
os.chdir(agent.config.workspace_path)
logger.info(
f"Executing command '{command_line}' in working directory '{os.getcwd()}'"
)
result = subprocess.run(command_line, capture_output=True, shell=True)
output = f"STDOUT:\n{result.stdout}\nSTDERR:\n{result.stderr}"
# Change back to whatever the prior working dir was
os.chdir(current_dir)
return output
@command(
"execute_shell_popen",
"Executes a Shell Command, non-interactive commands only",
{
"query": {
"type": "string",
"description": "The search query",
"required": True,
}
},
lambda config: config.execute_local_commands,
"You are not allowed to run local shell commands. To execute"
" shell commands, EXECUTE_LOCAL_COMMANDS must be set to 'True' "
"in your config. Do not attempt to bypass the restriction.",
)
def execute_shell_popen(command_line, agent: Agent) -> str:
"""Execute a shell command with Popen and returns an english description
of the event and the process id
Args:
command_line (str): The command line to execute
Returns:
str: Description of the fact that the process started and its id
"""
if not validate_command(command_line, agent.config):
logger.info(f"Command '{command_line}' not allowed")
return "Error: This Shell Command is not allowed."
current_dir = os.getcwd()
# Change dir into workspace if necessary
if agent.config.workspace_path not in current_dir:
os.chdir(agent.config.workspace_path)
logger.info(
f"Executing command '{command_line}' in working directory '{os.getcwd()}'"
)
do_not_show_output = subprocess.DEVNULL
process = subprocess.Popen(
command_line, shell=True, stdout=do_not_show_output, stderr=do_not_show_output
)
# Change back to whatever the prior working dir was
os.chdir(current_dir)
return f"Subprocess started with PID:'{str(process.pid)}'"
def we_are_running_in_a_docker_container() -> bool:
"""Check if we are running in a Docker container
Returns:
bool: True if we are running in a Docker container, False otherwise
"""
return os.path.exists("/.dockerenv")

View File

@@ -0,0 +1,340 @@
"""File operations for AutoGPT"""
from __future__ import annotations
import contextlib
import hashlib
import os
import os.path
from pathlib import Path
from typing import Generator, Literal
from autogpt.agents.agent import Agent
from autogpt.command_decorator import command
from autogpt.logs import logger
from autogpt.memory.vector import MemoryItem, VectorMemory
from .decorators import sanitize_path_arg
from .file_operations_utils import read_textual_file
Operation = Literal["write", "append", "delete"]
def text_checksum(text: str) -> str:
"""Get the hex checksum for the given text."""
return hashlib.md5(text.encode("utf-8")).hexdigest()
def operations_from_log(
log_path: str,
) -> Generator[tuple[Operation, str, str | None], None, None]:
"""Parse the file operations log and return a tuple containing the log entries"""
try:
log = open(log_path, "r", encoding="utf-8")
except FileNotFoundError:
return
for line in log:
line = line.replace("File Operation Logger", "").strip()
if not line:
continue
operation, tail = line.split(": ", maxsplit=1)
operation = operation.strip()
if operation in ("write", "append"):
try:
path, checksum = (x.strip() for x in tail.rsplit(" #", maxsplit=1))
except ValueError:
logger.warn(f"File log entry lacks checksum: '{line}'")
path, checksum = tail.strip(), None
yield (operation, path, checksum)
elif operation == "delete":
yield (operation, tail.strip(), None)
log.close()
def file_operations_state(log_path: str) -> dict[str, str]:
"""Iterates over the operations log and returns the expected state.
Parses a log file at config.file_logger_path to construct a dictionary that maps
each file path written or appended to its checksum. Deleted files are removed
from the dictionary.
Returns:
A dictionary mapping file paths to their checksums.
Raises:
FileNotFoundError: If config.file_logger_path is not found.
ValueError: If the log file content is not in the expected format.
"""
state = {}
for operation, path, checksum in operations_from_log(log_path):
if operation in ("write", "append"):
state[path] = checksum
elif operation == "delete":
del state[path]
return state
@sanitize_path_arg("filename")
def is_duplicate_operation(
operation: Operation, filename: str, agent: Agent, checksum: str | None = None
) -> bool:
"""Check if the operation has already been performed
Args:
operation: The operation to check for
filename: The name of the file to check for
agent: The agent
checksum: The checksum of the contents to be written
Returns:
True if the operation has already been performed on the file
"""
# Make the filename into a relative path if possible
with contextlib.suppress(ValueError):
filename = str(Path(filename).relative_to(agent.workspace.root))
state = file_operations_state(agent.config.file_logger_path)
if operation == "delete" and filename not in state:
return True
if operation == "write" and state.get(filename) == checksum:
return True
return False
@sanitize_path_arg("filename")
def log_operation(
operation: Operation, filename: str, agent: Agent, checksum: str | None = None
) -> None:
"""Log the file operation to the file_logger.txt
Args:
operation: The operation to log
filename: The name of the file the operation was performed on
checksum: The checksum of the contents to be written
"""
# Make the filename into a relative path if possible
with contextlib.suppress(ValueError):
filename = str(Path(filename).relative_to(agent.workspace.root))
log_entry = f"{operation}: {filename}"
if checksum is not None:
log_entry += f" #{checksum}"
logger.debug(f"Logging file operation: {log_entry}")
append_to_file(
agent.config.file_logger_path, f"{log_entry}\n", agent, should_log=False
)
@command(
"read_file",
"Read an existing file",
{
"filename": {
"type": "string",
"description": "The path of the file to read",
"required": True,
}
},
)
@sanitize_path_arg("filename")
def read_file(filename: str, agent: Agent) -> str:
"""Read a file and return the contents
Args:
filename (str): The name of the file to read
Returns:
str: The contents of the file
"""
try:
content = read_textual_file(filename, logger)
# TODO: invalidate/update memory when file is edited
file_memory = MemoryItem.from_text_file(content, filename, agent.config)
if len(file_memory.chunks) > 1:
return file_memory.summary
return content
except Exception as e:
return f"Error: {str(e)}"
def ingest_file(
filename: str,
memory: VectorMemory,
) -> None:
"""
Ingest a file by reading its content, splitting it into chunks with a specified
maximum length and overlap, and adding the chunks to the memory storage.
Args:
filename: The name of the file to ingest
memory: An object with an add() method to store the chunks in memory
"""
try:
logger.info(f"Ingesting file {filename}")
content = read_file(filename)
# TODO: differentiate between different types of files
file_memory = MemoryItem.from_text_file(content, filename)
logger.debug(f"Created memory: {file_memory.dump(True)}")
memory.add(file_memory)
logger.info(f"Ingested {len(file_memory.e_chunks)} chunks from {filename}")
except Exception as err:
logger.warn(f"Error while ingesting file '{filename}': {err}")
@command(
"write_to_file",
"Writes to a file",
{
"filename": {
"type": "string",
"description": "The name of the file to write to",
"required": True,
},
"text": {
"type": "string",
"description": "The text to write to the file",
"required": True,
},
},
aliases=["write_file", "create_file"],
)
@sanitize_path_arg("filename")
def write_to_file(filename: str, text: str, agent: Agent) -> str:
"""Write text to a file
Args:
filename (str): The name of the file to write to
text (str): The text to write to the file
Returns:
str: A message indicating success or failure
"""
checksum = text_checksum(text)
if is_duplicate_operation("write", filename, agent, checksum):
return "Error: File has already been updated."
try:
directory = os.path.dirname(filename)
os.makedirs(directory, exist_ok=True)
with open(filename, "w", encoding="utf-8") as f:
f.write(text)
log_operation("write", filename, agent, checksum)
return "File written to successfully."
except Exception as err:
return f"Error: {err}"
@command(
"append_to_file",
"Appends to a file",
{
"filename": {
"type": "string",
"description": "The name of the file to write to",
"required": True,
},
"text": {
"type": "string",
"description": "The text to write to the file",
"required": True,
},
},
)
@sanitize_path_arg("filename")
def append_to_file(
filename: str, text: str, agent: Agent, should_log: bool = True
) -> str:
"""Append text to a file
Args:
filename (str): The name of the file to append to
text (str): The text to append to the file
should_log (bool): Should log output
Returns:
str: A message indicating success or failure
"""
try:
directory = os.path.dirname(filename)
os.makedirs(directory, exist_ok=True)
with open(filename, "a", encoding="utf-8") as f:
f.write(text)
if should_log:
with open(filename, "r", encoding="utf-8") as f:
checksum = text_checksum(f.read())
log_operation("append", filename, agent, checksum=checksum)
return "Text appended successfully."
except Exception as err:
return f"Error: {err}"
@command(
"delete_file",
"Deletes a file",
{
"filename": {
"type": "string",
"description": "The name of the file to delete",
"required": True,
}
},
)
@sanitize_path_arg("filename")
def delete_file(filename: str, agent: Agent) -> str:
"""Delete a file
Args:
filename (str): The name of the file to delete
Returns:
str: A message indicating success or failure
"""
if is_duplicate_operation("delete", filename, agent):
return "Error: File has already been deleted."
try:
os.remove(filename)
log_operation("delete", filename, agent)
return "File deleted successfully."
except Exception as err:
return f"Error: {err}"
@command(
"list_files",
"Lists Files in a Directory",
{
"directory": {
"type": "string",
"description": "The directory to list files in",
"required": True,
}
},
)
@sanitize_path_arg("directory")
def list_files(directory: str, agent: Agent) -> list[str]:
"""lists files in a directory recursively
Args:
directory (str): The directory to search in
Returns:
list[str]: A list of files found in the directory
"""
found_files = []
for root, _, files in os.walk(directory):
for file in files:
if file.startswith("."):
continue
relative_path = os.path.relpath(
os.path.join(root, file), agent.config.workspace_path
)
found_files.append(relative_path)
return found_files

View File

@@ -0,0 +1,161 @@
import json
import os
import charset_normalizer
import docx
import markdown
import PyPDF2
import yaml
from bs4 import BeautifulSoup
from pylatexenc.latex2text import LatexNodes2Text
from autogpt import logs
from autogpt.logs import logger
class ParserStrategy:
def read(self, file_path: str) -> str:
raise NotImplementedError
# Basic text file reading
class TXTParser(ParserStrategy):
def read(self, file_path: str) -> str:
charset_match = charset_normalizer.from_path(file_path).best()
logger.debug(f"Reading '{file_path}' with encoding '{charset_match.encoding}'")
return str(charset_match)
# Reading text from binary file using pdf parser
class PDFParser(ParserStrategy):
def read(self, file_path: str) -> str:
parser = PyPDF2.PdfReader(file_path)
text = ""
for page_idx in range(len(parser.pages)):
text += parser.pages[page_idx].extract_text()
return text
# Reading text from binary file using docs parser
class DOCXParser(ParserStrategy):
def read(self, file_path: str) -> str:
doc_file = docx.Document(file_path)
text = ""
for para in doc_file.paragraphs:
text += para.text
return text
# Reading as dictionary and returning string format
class JSONParser(ParserStrategy):
def read(self, file_path: str) -> str:
with open(file_path, "r") as f:
data = json.load(f)
text = str(data)
return text
class XMLParser(ParserStrategy):
def read(self, file_path: str) -> str:
with open(file_path, "r") as f:
soup = BeautifulSoup(f, "xml")
text = soup.get_text()
return text
# Reading as dictionary and returning string format
class YAMLParser(ParserStrategy):
def read(self, file_path: str) -> str:
with open(file_path, "r") as f:
data = yaml.load(f, Loader=yaml.FullLoader)
text = str(data)
return text
class HTMLParser(ParserStrategy):
def read(self, file_path: str) -> str:
with open(file_path, "r") as f:
soup = BeautifulSoup(f, "html.parser")
text = soup.get_text()
return text
class MarkdownParser(ParserStrategy):
def read(self, file_path: str) -> str:
with open(file_path, "r") as f:
html = markdown.markdown(f.read())
text = "".join(BeautifulSoup(html, "html.parser").findAll(string=True))
return text
class LaTeXParser(ParserStrategy):
def read(self, file_path: str) -> str:
with open(file_path, "r") as f:
latex = f.read()
text = LatexNodes2Text().latex_to_text(latex)
return text
class FileContext:
def __init__(self, parser: ParserStrategy, logger: logs.Logger):
self.parser = parser
self.logger = logger
def set_parser(self, parser: ParserStrategy) -> None:
self.logger.debug(f"Setting Context Parser to {parser}")
self.parser = parser
def read_file(self, file_path) -> str:
self.logger.debug(f"Reading file {file_path} with parser {self.parser}")
return self.parser.read(file_path)
extension_to_parser = {
".txt": TXTParser(),
".csv": TXTParser(),
".pdf": PDFParser(),
".docx": DOCXParser(),
".json": JSONParser(),
".xml": XMLParser(),
".yaml": YAMLParser(),
".yml": YAMLParser(),
".html": HTMLParser(),
".htm": HTMLParser(),
".xhtml": HTMLParser(),
".md": MarkdownParser(),
".markdown": MarkdownParser(),
".tex": LaTeXParser(),
}
def is_file_binary_fn(file_path: str):
"""Given a file path load all its content and checks if the null bytes is present
Args:
file_path (_type_): _description_
Returns:
bool: is_binary
"""
with open(file_path, "rb") as f:
file_data = f.read()
if b"\x00" in file_data:
return True
return False
def read_textual_file(file_path: str, logger: logs.Logger) -> str:
if not os.path.isfile(file_path):
raise FileNotFoundError(
f"read_file {file_path} failed: no such file or directory"
)
is_binary = is_file_binary_fn(file_path)
file_extension = os.path.splitext(file_path)[1].lower()
parser = extension_to_parser.get(file_extension)
if not parser:
if is_binary:
raise ValueError(f"Unsupported binary file format: {file_extension}")
# fallback to txt file parser (to support script and code files loading)
parser = TXTParser()
file_context = FileContext(parser, logger)
return file_context.read_file(file_path)

View File

@@ -0,0 +1,52 @@
"""Git operations for autogpt"""
from git.repo import Repo
from autogpt.agents.agent import Agent
from autogpt.command_decorator import command
from autogpt.url_utils.validators import validate_url
from .decorators import sanitize_path_arg
@command(
"clone_repository",
"Clones a Repository",
{
"url": {
"type": "string",
"description": "The URL of the repository to clone",
"required": True,
},
"clone_path": {
"type": "string",
"description": "The path to clone the repository to",
"required": True,
},
},
lambda config: bool(config.github_username and config.github_api_key),
"Configure github_username and github_api_key.",
)
@sanitize_path_arg("clone_path")
@validate_url
def clone_repository(url: str, clone_path: str, agent: Agent) -> str:
"""Clone a GitHub repository locally.
Args:
url (str): The URL of the repository to clone.
clone_path (str): The path to clone the repository to.
Returns:
str: The result of the clone operation.
"""
split_url = url.split("//")
auth_repo_url = (
f"//{agent.config.github_username}:{agent.config.github_api_key}@".join(
split_url
)
)
try:
Repo.clone_from(url=auth_repo_url, to_path=clone_path)
return f"""Cloned {url} to {clone_path}"""
except Exception as e:
return f"Error: {str(e)}"

View File

@@ -0,0 +1,200 @@
""" Image Generation Module for AutoGPT."""
import io
import json
import time
import uuid
from base64 import b64decode
import openai
import requests
from PIL import Image
from autogpt.agents.agent import Agent
from autogpt.command_decorator import command
from autogpt.logs import logger
@command(
"generate_image",
"Generates an Image",
{
"prompt": {
"type": "string",
"description": "The prompt used to generate the image",
"required": True,
},
},
lambda config: bool(config.image_provider),
"Requires a image provider to be set.",
)
def generate_image(prompt: str, agent: Agent, size: int = 256) -> str:
"""Generate an image from a prompt.
Args:
prompt (str): The prompt to use
size (int, optional): The size of the image. Defaults to 256. (Not supported by HuggingFace)
Returns:
str: The filename of the image
"""
filename = agent.config.workspace_path / f"{str(uuid.uuid4())}.jpg"
# DALL-E
if agent.config.image_provider == "dalle":
return generate_image_with_dalle(prompt, filename, size, agent)
# HuggingFace
elif agent.config.image_provider == "huggingface":
return generate_image_with_hf(prompt, filename, agent)
# SD WebUI
elif agent.config.image_provider == "sdwebui":
return generate_image_with_sd_webui(prompt, filename, agent, size)
return "No Image Provider Set"
def generate_image_with_hf(prompt: str, filename: str, agent: Agent) -> str:
"""Generate an image with HuggingFace's API.
Args:
prompt (str): The prompt to use
filename (str): The filename to save the image to
Returns:
str: The filename of the image
"""
API_URL = f"https://api-inference.huggingface.co/models/{agent.config.huggingface_image_model}"
if agent.config.huggingface_api_token is None:
raise ValueError(
"You need to set your Hugging Face API token in the config file."
)
headers = {
"Authorization": f"Bearer {agent.config.huggingface_api_token}",
"X-Use-Cache": "false",
}
retry_count = 0
while retry_count < 10:
response = requests.post(
API_URL,
headers=headers,
json={
"inputs": prompt,
},
)
if response.ok:
try:
image = Image.open(io.BytesIO(response.content))
logger.info(f"Image Generated for prompt:{prompt}")
image.save(filename)
return f"Saved to disk:{filename}"
except Exception as e:
logger.error(e)
break
else:
try:
error = json.loads(response.text)
if "estimated_time" in error:
delay = error["estimated_time"]
logger.debug(response.text)
logger.info("Retrying in", delay)
time.sleep(delay)
else:
break
except Exception as e:
logger.error(e)
break
retry_count += 1
return f"Error creating image."
def generate_image_with_dalle(
prompt: str, filename: str, size: int, agent: Agent
) -> str:
"""Generate an image with DALL-E.
Args:
prompt (str): The prompt to use
filename (str): The filename to save the image to
size (int): The size of the image
Returns:
str: The filename of the image
"""
# Check for supported image sizes
if size not in [256, 512, 1024]:
closest = min([256, 512, 1024], key=lambda x: abs(x - size))
logger.info(
f"DALL-E only supports image sizes of 256x256, 512x512, or 1024x1024. Setting to {closest}, was {size}."
)
size = closest
response = openai.Image.create(
prompt=prompt,
n=1,
size=f"{size}x{size}",
response_format="b64_json",
api_key=agent.config.openai_api_key,
)
logger.info(f"Image Generated for prompt:{prompt}")
image_data = b64decode(response["data"][0]["b64_json"])
with open(filename, mode="wb") as png:
png.write(image_data)
return f"Saved to disk:{filename}"
def generate_image_with_sd_webui(
prompt: str,
filename: str,
agent: Agent,
size: int = 512,
negative_prompt: str = "",
extra: dict = {},
) -> str:
"""Generate an image with Stable Diffusion webui.
Args:
prompt (str): The prompt to use
filename (str): The filename to save the image to
size (int, optional): The size of the image. Defaults to 256.
negative_prompt (str, optional): The negative prompt to use. Defaults to "".
extra (dict, optional): Extra parameters to pass to the API. Defaults to {}.
Returns:
str: The filename of the image
"""
# Create a session and set the basic auth if needed
s = requests.Session()
if agent.config.sd_webui_auth:
username, password = agent.config.sd_webui_auth.split(":")
s.auth = (username, password or "")
# Generate the images
response = requests.post(
f"{agent.config.sd_webui_url}/sdapi/v1/txt2img",
json={
"prompt": prompt,
"negative_prompt": negative_prompt,
"sampler_index": "DDIM",
"steps": 20,
"config_scale": 7.0,
"width": size,
"height": size,
"n_iter": 1,
**extra,
},
)
logger.info(f"Image Generated for prompt:{prompt}")
# Save the image to disk
response = response.json()
b64 = b64decode(response["images"][0].split(",", 1)[0])
image = Image.open(io.BytesIO(b64))
image.save(filename)
return f"Saved to disk:{filename}"

View File

@@ -0,0 +1,33 @@
"""Task Statuses module."""
from __future__ import annotations
from typing import NoReturn
from autogpt.agents.agent import Agent
from autogpt.command_decorator import command
from autogpt.logs import logger
@command(
"goals_accomplished",
"Goals are accomplished and there is nothing left to do",
{
"reason": {
"type": "string",
"description": "A summary to the user of how the goals were accomplished",
"required": True,
}
},
)
def task_complete(reason: str, agent: Agent) -> NoReturn:
"""
A function that takes in a string and exits the program
Parameters:
reason (str): A summary to the user of how the goals were accomplished.
Returns:
A result string from create chat completion. A list of suggestions to
improve the code.
"""
logger.info(title="Shutting down...\n", message=reason)
quit()

10
autogpt/commands/times.py Normal file
View File

@@ -0,0 +1,10 @@
from datetime import datetime
def get_datetime() -> str:
"""Return the current date and time
Returns:
str: The current date and time
"""
return "Current date and time: " + datetime.now().strftime("%Y-%m-%d %H:%M:%S")

View File

@@ -0,0 +1,143 @@
"""Google search command for Autogpt."""
from __future__ import annotations
import json
import time
from itertools import islice
from duckduckgo_search import DDGS
from autogpt.agents.agent import Agent
from autogpt.command_decorator import command
DUCKDUCKGO_MAX_ATTEMPTS = 3
@command(
"web_search",
"Searches the web",
{
"query": {
"type": "string",
"description": "The search query",
"required": True,
}
},
aliases=["search"],
)
def web_search(query: str, agent: Agent, num_results: int = 8) -> str:
"""Return the results of a Google search
Args:
query (str): The search query.
num_results (int): The number of results to return.
Returns:
str: The results of the search.
"""
search_results = []
attempts = 0
while attempts < DUCKDUCKGO_MAX_ATTEMPTS:
if not query:
return json.dumps(search_results)
results = DDGS().text(query)
search_results = list(islice(results, num_results))
if search_results:
break
time.sleep(1)
attempts += 1
results = json.dumps(search_results, ensure_ascii=False, indent=4)
return safe_google_results(results)
@command(
"google",
"Google Search",
{
"query": {
"type": "string",
"description": "The search query",
"required": True,
}
},
lambda config: bool(config.google_api_key)
and bool(config.google_custom_search_engine_id),
"Configure google_api_key and custom_search_engine_id.",
aliases=["search"],
)
def google(query: str, agent: Agent, num_results: int = 8) -> str | list[str]:
"""Return the results of a Google search using the official Google API
Args:
query (str): The search query.
num_results (int): The number of results to return.
Returns:
str: The results of the search.
"""
from googleapiclient.discovery import build
from googleapiclient.errors import HttpError
try:
# Get the Google API key and Custom Search Engine ID from the config file
api_key = agent.config.google_api_key
custom_search_engine_id = agent.config.google_custom_search_engine_id
# Initialize the Custom Search API service
service = build("customsearch", "v1", developerKey=api_key)
# Send the search query and retrieve the results
result = (
service.cse()
.list(q=query, cx=custom_search_engine_id, num=num_results)
.execute()
)
# Extract the search result items from the response
search_results = result.get("items", [])
# Create a list of only the URLs from the search results
search_results_links = [item["link"] for item in search_results]
except HttpError as e:
# Handle errors in the API call
error_details = json.loads(e.content.decode())
# Check if the error is related to an invalid or missing API key
if error_details.get("error", {}).get(
"code"
) == 403 and "invalid API key" in error_details.get("error", {}).get(
"message", ""
):
return "Error: The provided Google API key is invalid or missing."
else:
return f"Error: {e}"
# google_result can be a list or a string depending on the search results
# Return the list of search result URLs
return safe_google_results(search_results_links)
def safe_google_results(results: str | list) -> str:
"""
Return the results of a Google search in a safe format.
Args:
results (str | list): The search results.
Returns:
str: The results of the search.
"""
if isinstance(results, list):
safe_message = json.dumps(
[result.encode("utf-8", "ignore").decode("utf-8") for result in results]
)
else:
safe_message = results.encode("utf-8", "ignore").decode("utf-8")
return safe_message

View File

@@ -0,0 +1,237 @@
"""Selenium web scraping module."""
from __future__ import annotations
import logging
from pathlib import Path
from sys import platform
from typing import Optional, Type
from bs4 import BeautifulSoup
from selenium.common.exceptions import WebDriverException
from selenium.webdriver.chrome.options import Options as ChromeOptions
from selenium.webdriver.chrome.service import Service as ChromeDriverService
from selenium.webdriver.chrome.webdriver import WebDriver as ChromeDriver
from selenium.webdriver.common.by import By
from selenium.webdriver.edge.options import Options as EdgeOptions
from selenium.webdriver.edge.service import Service as EdgeDriverService
from selenium.webdriver.edge.webdriver import WebDriver as EdgeDriver
from selenium.webdriver.firefox.options import Options as FirefoxOptions
from selenium.webdriver.firefox.service import Service as GeckoDriverService
from selenium.webdriver.firefox.webdriver import WebDriver as FirefoxDriver
from selenium.webdriver.remote.webdriver import WebDriver
from selenium.webdriver.safari.options import Options as SafariOptions
from selenium.webdriver.safari.webdriver import WebDriver as SafariDriver
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.support.wait import WebDriverWait
from webdriver_manager.chrome import ChromeDriverManager
from webdriver_manager.firefox import GeckoDriverManager
from webdriver_manager.microsoft import EdgeChromiumDriverManager as EdgeDriverManager
from autogpt.agents.agent import Agent
from autogpt.command_decorator import command
from autogpt.logs import logger
from autogpt.memory.vector import MemoryItem, get_memory
from autogpt.processing.html import extract_hyperlinks, format_hyperlinks
from autogpt.url_utils.validators import validate_url
BrowserOptions = ChromeOptions | EdgeOptions | FirefoxOptions | SafariOptions
FILE_DIR = Path(__file__).parent.parent
@command(
"browse_website",
"Browses a Website",
{
"url": {"type": "string", "description": "The URL to visit", "required": True},
"question": {
"type": "string",
"description": "What you want to find on the website",
"required": True,
},
},
)
@validate_url
def browse_website(url: str, question: str, agent: Agent) -> str:
"""Browse a website and return the answer and links to the user
Args:
url (str): The url of the website to browse
question (str): The question asked by the user
Returns:
Tuple[str, WebDriver]: The answer and links to the user and the webdriver
"""
try:
driver, text = scrape_text_with_selenium(url, agent)
except WebDriverException as e:
# These errors are often quite long and include lots of context.
# Just grab the first line.
msg = e.msg.split("\n")[0]
return f"Error: {msg}"
add_header(driver)
summary = summarize_memorize_webpage(url, text, question, agent, driver)
links = scrape_links_with_selenium(driver, url)
# Limit links to 5
if len(links) > 5:
links = links[:5]
close_browser(driver)
return f"Answer gathered from website: {summary}\n\nLinks: {links}"
def scrape_text_with_selenium(url: str, agent: Agent) -> tuple[WebDriver, str]:
"""Scrape text from a website using selenium
Args:
url (str): The url of the website to scrape
Returns:
Tuple[WebDriver, str]: The webdriver and the text scraped from the website
"""
logging.getLogger("selenium").setLevel(logging.CRITICAL)
options_available: dict[str, Type[BrowserOptions]] = {
"chrome": ChromeOptions,
"edge": EdgeOptions,
"firefox": FirefoxOptions,
"safari": SafariOptions,
}
options: BrowserOptions = options_available[agent.config.selenium_web_browser]()
options.add_argument(
"user-agent=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/112.0.5615.49 Safari/537.36"
)
if agent.config.selenium_web_browser == "firefox":
if agent.config.selenium_headless:
options.headless = True
options.add_argument("--disable-gpu")
driver = FirefoxDriver(
service=GeckoDriverService(GeckoDriverManager().install()), options=options
)
elif agent.config.selenium_web_browser == "edge":
driver = EdgeDriver(
service=EdgeDriverService(EdgeDriverManager().install()), options=options
)
elif agent.config.selenium_web_browser == "safari":
# Requires a bit more setup on the users end
# See https://developer.apple.com/documentation/webkit/testing_with_webdriver_in_safari
driver = SafariDriver(options=options)
else:
if platform == "linux" or platform == "linux2":
options.add_argument("--disable-dev-shm-usage")
options.add_argument("--remote-debugging-port=9222")
options.add_argument("--no-sandbox")
if agent.config.selenium_headless:
options.add_argument("--headless=new")
options.add_argument("--disable-gpu")
chromium_driver_path = Path("/usr/bin/chromedriver")
driver = ChromeDriver(
service=ChromeDriverService(str(chromium_driver_path))
if chromium_driver_path.exists()
else ChromeDriverService(ChromeDriverManager().install()),
options=options,
)
driver.get(url)
WebDriverWait(driver, 10).until(
EC.presence_of_element_located((By.TAG_NAME, "body"))
)
# Get the HTML content directly from the browser's DOM
page_source = driver.execute_script("return document.body.outerHTML;")
soup = BeautifulSoup(page_source, "html.parser")
for script in soup(["script", "style"]):
script.extract()
text = soup.get_text()
lines = (line.strip() for line in text.splitlines())
chunks = (phrase.strip() for line in lines for phrase in line.split(" "))
text = "\n".join(chunk for chunk in chunks if chunk)
return driver, text
def scrape_links_with_selenium(driver: WebDriver, url: str) -> list[str]:
"""Scrape links from a website using selenium
Args:
driver (WebDriver): The webdriver to use to scrape the links
Returns:
List[str]: The links scraped from the website
"""
page_source = driver.page_source
soup = BeautifulSoup(page_source, "html.parser")
for script in soup(["script", "style"]):
script.extract()
hyperlinks = extract_hyperlinks(soup, url)
return format_hyperlinks(hyperlinks)
def close_browser(driver: WebDriver) -> None:
"""Close the browser
Args:
driver (WebDriver): The webdriver to close
Returns:
None
"""
driver.quit()
def add_header(driver: WebDriver) -> None:
"""Add a header to the website
Args:
driver (WebDriver): The webdriver to use to add the header
Returns:
None
"""
try:
with open(f"{FILE_DIR}/js/overlay.js", "r") as overlay_file:
overlay_script = overlay_file.read()
driver.execute_script(overlay_script)
except Exception as e:
print(f"Error executing overlay.js: {e}")
def summarize_memorize_webpage(
url: str,
text: str,
question: str,
agent: Agent,
driver: Optional[WebDriver] = None,
) -> str:
"""Summarize text using the OpenAI API
Args:
url (str): The url of the text
text (str): The text to summarize
question (str): The question to ask the model
driver (WebDriver): The webdriver to use to scroll the page
Returns:
str: The summary of the text
"""
if not text:
return "Error: No text to summarize"
text_length = len(text)
logger.info(f"Text length: {text_length} characters")
memory = get_memory(agent.config)
new_memory = MemoryItem.from_webpage(text, url, agent.config, question=question)
memory.add(new_memory)
return new_memory.summary

View File

@@ -0,0 +1,12 @@
"""
This module contains the configuration classes for AutoGPT.
"""
from .ai_config import AIConfig
from .config import Config, ConfigBuilder, check_openai_api_key
__all__ = [
"check_openai_api_key",
"AIConfig",
"Config",
"ConfigBuilder",
]

159
autogpt/config/ai_config.py Normal file
View File

@@ -0,0 +1,159 @@
# sourcery skip: do-not-use-staticmethod
"""
A module that contains the AIConfig class object that contains the configuration
"""
from __future__ import annotations
import platform
from pathlib import Path
from typing import TYPE_CHECKING, Optional
import distro
import yaml
if TYPE_CHECKING:
from autogpt.models.command_registry import CommandRegistry
from autogpt.prompts.generator import PromptGenerator
class AIConfig:
"""
A class object that contains the configuration information for the AI
Attributes:
ai_name (str): The name of the AI.
ai_role (str): The description of the AI's role.
ai_goals (list): The list of objectives the AI is supposed to complete.
api_budget (float): The maximum dollar value for API calls (0.0 means infinite)
"""
def __init__(
self,
ai_name: str = "",
ai_role: str = "",
ai_goals: list[str] = [],
api_budget: float = 0.0,
) -> None:
"""
Initialize a class instance
Parameters:
ai_name (str): The name of the AI.
ai_role (str): The description of the AI's role.
ai_goals (list): The list of objectives the AI is supposed to complete.
api_budget (float): The maximum dollar value for API calls (0.0 means infinite)
Returns:
None
"""
self.ai_name = ai_name
self.ai_role = ai_role
self.ai_goals = ai_goals
self.api_budget = api_budget
self.prompt_generator: PromptGenerator | None = None
self.command_registry: CommandRegistry | None = None
@staticmethod
def load(ai_settings_file: str | Path) -> "AIConfig":
"""
Returns class object with parameters (ai_name, ai_role, ai_goals, api_budget)
loaded from yaml file if yaml file exists, else returns class with no parameters.
Parameters:
ai_settings_file (Path): The path to the config yaml file.
Returns:
cls (object): An instance of given cls object
"""
try:
with open(ai_settings_file, encoding="utf-8") as file:
config_params = yaml.load(file, Loader=yaml.FullLoader) or {}
except FileNotFoundError:
config_params = {}
ai_name = config_params.get("ai_name", "")
ai_role = config_params.get("ai_role", "")
ai_goals = [
str(goal).strip("{}").replace("'", "").replace('"', "")
if isinstance(goal, dict)
else str(goal)
for goal in config_params.get("ai_goals", [])
]
api_budget = config_params.get("api_budget", 0.0)
return AIConfig(ai_name, ai_role, ai_goals, api_budget)
def save(self, ai_settings_file: str | Path) -> None:
"""
Saves the class parameters to the specified file yaml file path as a yaml file.
Parameters:
ai_settings_file (Path): The path to the config yaml file.
Returns:
None
"""
config = {
"ai_name": self.ai_name,
"ai_role": self.ai_role,
"ai_goals": self.ai_goals,
"api_budget": self.api_budget,
}
with open(ai_settings_file, "w", encoding="utf-8") as file:
yaml.dump(config, file, allow_unicode=True)
def construct_full_prompt(
self, config, prompt_generator: Optional[PromptGenerator] = None
) -> str:
"""
Returns a prompt to the user with the class information in an organized fashion.
Parameters:
None
Returns:
full_prompt (str): A string containing the initial prompt for the user
including the ai_name, ai_role, ai_goals, and api_budget.
"""
prompt_start = (
"Your decisions must always be made independently without"
" seeking user assistance. Play to your strengths as an LLM and pursue"
" simple strategies with no legal complications."
""
)
from autogpt.prompts.prompt import build_default_prompt_generator
if prompt_generator is None:
prompt_generator = build_default_prompt_generator(config)
prompt_generator.goals = self.ai_goals
prompt_generator.name = self.ai_name
prompt_generator.role = self.ai_role
prompt_generator.command_registry = self.command_registry
for plugin in config.plugins:
if not plugin.can_handle_post_prompt():
continue
prompt_generator = plugin.post_prompt(prompt_generator)
if config.execute_local_commands:
# add OS info to prompt
os_name = platform.system()
os_info = (
platform.platform(terse=True)
if os_name != "Linux"
else distro.name(pretty=True)
)
prompt_start += f"\nThe OS you are running on is: {os_info}"
# Construct full prompt
full_prompt = f"You are {prompt_generator.name}, {prompt_generator.role}\n{prompt_start}\n\nGOALS:\n\n"
for i, goal in enumerate(self.ai_goals):
full_prompt += f"{i+1}. {goal}\n"
if self.api_budget > 0.0:
full_prompt += f"\nIt takes money to let you run. Your API budget is ${self.api_budget:.3f}"
self.prompt_generator = prompt_generator
full_prompt += f"\n\n{prompt_generator.generate_prompt_string(config)}"
return full_prompt

399
autogpt/config/config.py Normal file
View File

@@ -0,0 +1,399 @@
"""Configuration class to store the state of bools for different scripts access."""
from __future__ import annotations
import contextlib
import os
import re
from pathlib import Path
from typing import Any, Dict, Optional, Union
import yaml
from auto_gpt_plugin_template import AutoGPTPluginTemplate
from colorama import Fore
from pydantic import Field, validator
from autogpt.core.configuration.schema import Configurable, SystemSettings
from autogpt.plugins.plugins_config import PluginsConfig
AI_SETTINGS_FILE = "ai_settings.yaml"
AZURE_CONFIG_FILE = "azure.yaml"
PLUGINS_CONFIG_FILE = "plugins_config.yaml"
PROMPT_SETTINGS_FILE = "prompt_settings.yaml"
GPT_4_MODEL = "gpt-4"
GPT_3_MODEL = "gpt-3.5-turbo"
class Config(SystemSettings, arbitrary_types_allowed=True):
name: str = "Auto-GPT configuration"
description: str = "Default configuration for the Auto-GPT application."
########################
# Application Settings #
########################
skip_news: bool = False
skip_reprompt: bool = False
authorise_key: str = "y"
exit_key: str = "n"
debug_mode: bool = False
plain_output: bool = False
chat_messages_enabled: bool = True
# TTS configuration
speak_mode: bool = False
text_to_speech_provider: str = "gtts"
streamelements_voice: str = "Brian"
elevenlabs_voice_id: Optional[str] = None
##########################
# Agent Control Settings #
##########################
# Paths
ai_settings_file: str = AI_SETTINGS_FILE
prompt_settings_file: str = PROMPT_SETTINGS_FILE
workdir: Path = None
workspace_path: Optional[Path] = None
file_logger_path: Optional[str] = None
# Model configuration
fast_llm: str = "gpt-3.5-turbo"
smart_llm: str = "gpt-4"
temperature: float = 0
openai_functions: bool = False
embedding_model: str = "text-embedding-ada-002"
browse_spacy_language_model: str = "en_core_web_sm"
# Run loop configuration
continuous_mode: bool = False
continuous_limit: int = 0
##########
# Memory #
##########
memory_backend: str = "json_file"
memory_index: str = "auto-gpt-memory"
redis_host: str = "localhost"
redis_port: int = 6379
redis_password: str = ""
wipe_redis_on_start: bool = True
############
# Commands #
############
# General
disabled_command_categories: list[str] = Field(default_factory=list)
# File ops
restrict_to_workspace: bool = True
allow_downloads: bool = False
# Shell commands
shell_command_control: str = "denylist"
execute_local_commands: bool = False
shell_denylist: list[str] = Field(default_factory=lambda: ["sudo", "su"])
shell_allowlist: list[str] = Field(default_factory=list)
# Text to image
image_provider: Optional[str] = None
huggingface_image_model: str = "CompVis/stable-diffusion-v1-4"
sd_webui_url: Optional[str] = "http://localhost:7860"
image_size: int = 256
# Audio to text
audio_to_text_provider: str = "huggingface"
huggingface_audio_to_text_model: Optional[str] = None
# Web browsing
selenium_web_browser: str = "chrome"
selenium_headless: bool = True
user_agent: str = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36"
###################
# Plugin Settings #
###################
plugins_dir: str = "plugins"
plugins_config_file: str = PLUGINS_CONFIG_FILE
plugins_config: PluginsConfig = Field(
default_factory=lambda: PluginsConfig(plugins={})
)
plugins: list[AutoGPTPluginTemplate] = Field(default_factory=list, exclude=True)
plugins_allowlist: list[str] = Field(default_factory=list)
plugins_denylist: list[str] = Field(default_factory=list)
plugins_openai: list[str] = Field(default_factory=list)
###############
# Credentials #
###############
# OpenAI
openai_api_key: Optional[str] = None
openai_api_type: Optional[str] = None
openai_api_base: Optional[str] = None
openai_api_version: Optional[str] = None
openai_organization: Optional[str] = None
use_azure: bool = False
azure_config_file: Optional[str] = AZURE_CONFIG_FILE
azure_model_to_deployment_id_map: Optional[Dict[str, str]] = None
# Elevenlabs
elevenlabs_api_key: Optional[str] = None
# Github
github_api_key: Optional[str] = None
github_username: Optional[str] = None
# Google
google_api_key: Optional[str] = None
google_custom_search_engine_id: Optional[str] = None
# Huggingface
huggingface_api_token: Optional[str] = None
# Stable Diffusion
sd_webui_auth: Optional[str] = None
@validator("plugins", each_item=True)
def validate_plugins(cls, p: AutoGPTPluginTemplate | Any):
assert issubclass(
p.__class__, AutoGPTPluginTemplate
), f"{p} does not subclass AutoGPTPluginTemplate"
assert (
p.__class__.__name__ != "AutoGPTPluginTemplate"
), f"Plugins must subclass AutoGPTPluginTemplate; {p} is a template instance"
return p
def get_openai_credentials(self, model: str) -> dict[str, str]:
credentials = {
"api_key": self.openai_api_key,
"api_base": self.openai_api_base,
"organization": self.openai_organization,
}
if self.use_azure:
azure_credentials = self.get_azure_credentials(model)
credentials.update(azure_credentials)
return credentials
def get_azure_credentials(self, model: str) -> dict[str, str]:
"""Get the kwargs for the Azure API."""
# Fix --gpt3only and --gpt4only in combination with Azure
fast_llm = (
self.fast_llm
if not (
self.fast_llm == self.smart_llm
and self.fast_llm.startswith(GPT_4_MODEL)
)
else f"not_{self.fast_llm}"
)
smart_llm = (
self.smart_llm
if not (
self.smart_llm == self.fast_llm
and self.smart_llm.startswith(GPT_3_MODEL)
)
else f"not_{self.smart_llm}"
)
deployment_id = {
fast_llm: self.azure_model_to_deployment_id_map.get(
"fast_llm_deployment_id",
self.azure_model_to_deployment_id_map.get(
"fast_llm_model_deployment_id" # backwards compatibility
),
),
smart_llm: self.azure_model_to_deployment_id_map.get(
"smart_llm_deployment_id",
self.azure_model_to_deployment_id_map.get(
"smart_llm_model_deployment_id" # backwards compatibility
),
),
self.embedding_model: self.azure_model_to_deployment_id_map.get(
"embedding_model_deployment_id"
),
}.get(model, None)
kwargs = {
"api_type": self.openai_api_type,
"api_base": self.openai_api_base,
"api_version": self.openai_api_version,
}
if model == self.embedding_model:
kwargs["engine"] = deployment_id
else:
kwargs["deployment_id"] = deployment_id
return kwargs
class ConfigBuilder(Configurable[Config]):
default_settings = Config()
@classmethod
def build_config_from_env(cls, workdir: Path) -> Config:
"""Initialize the Config class"""
config_dict = {
"workdir": workdir,
"authorise_key": os.getenv("AUTHORISE_COMMAND_KEY"),
"exit_key": os.getenv("EXIT_KEY"),
"plain_output": os.getenv("PLAIN_OUTPUT", "False") == "True",
"shell_command_control": os.getenv("SHELL_COMMAND_CONTROL"),
"ai_settings_file": os.getenv("AI_SETTINGS_FILE", AI_SETTINGS_FILE),
"prompt_settings_file": os.getenv(
"PROMPT_SETTINGS_FILE", PROMPT_SETTINGS_FILE
),
"fast_llm": os.getenv("FAST_LLM", os.getenv("FAST_LLM_MODEL")),
"smart_llm": os.getenv("SMART_LLM", os.getenv("SMART_LLM_MODEL")),
"embedding_model": os.getenv("EMBEDDING_MODEL"),
"browse_spacy_language_model": os.getenv("BROWSE_SPACY_LANGUAGE_MODEL"),
"openai_api_key": os.getenv("OPENAI_API_KEY"),
"use_azure": os.getenv("USE_AZURE") == "True",
"azure_config_file": os.getenv("AZURE_CONFIG_FILE", AZURE_CONFIG_FILE),
"execute_local_commands": os.getenv("EXECUTE_LOCAL_COMMANDS", "False")
== "True",
"restrict_to_workspace": os.getenv("RESTRICT_TO_WORKSPACE", "True")
== "True",
"openai_functions": os.getenv("OPENAI_FUNCTIONS", "False") == "True",
"elevenlabs_api_key": os.getenv("ELEVENLABS_API_KEY"),
"streamelements_voice": os.getenv("STREAMELEMENTS_VOICE"),
"text_to_speech_provider": os.getenv("TEXT_TO_SPEECH_PROVIDER"),
"github_api_key": os.getenv("GITHUB_API_KEY"),
"github_username": os.getenv("GITHUB_USERNAME"),
"google_api_key": os.getenv("GOOGLE_API_KEY"),
"image_provider": os.getenv("IMAGE_PROVIDER"),
"huggingface_api_token": os.getenv("HUGGINGFACE_API_TOKEN"),
"huggingface_image_model": os.getenv("HUGGINGFACE_IMAGE_MODEL"),
"audio_to_text_provider": os.getenv("AUDIO_TO_TEXT_PROVIDER"),
"huggingface_audio_to_text_model": os.getenv(
"HUGGINGFACE_AUDIO_TO_TEXT_MODEL"
),
"sd_webui_url": os.getenv("SD_WEBUI_URL"),
"sd_webui_auth": os.getenv("SD_WEBUI_AUTH"),
"selenium_web_browser": os.getenv("USE_WEB_BROWSER"),
"selenium_headless": os.getenv("HEADLESS_BROWSER", "True") == "True",
"user_agent": os.getenv("USER_AGENT"),
"memory_backend": os.getenv("MEMORY_BACKEND"),
"memory_index": os.getenv("MEMORY_INDEX"),
"redis_host": os.getenv("REDIS_HOST"),
"redis_password": os.getenv("REDIS_PASSWORD"),
"wipe_redis_on_start": os.getenv("WIPE_REDIS_ON_START", "True") == "True",
"plugins_dir": os.getenv("PLUGINS_DIR"),
"plugins_config_file": os.getenv(
"PLUGINS_CONFIG_FILE", PLUGINS_CONFIG_FILE
),
"chat_messages_enabled": os.getenv("CHAT_MESSAGES_ENABLED") == "True",
}
config_dict["disabled_command_categories"] = _safe_split(
os.getenv("DISABLED_COMMAND_CATEGORIES")
)
config_dict["shell_denylist"] = _safe_split(
os.getenv("SHELL_DENYLIST", os.getenv("DENY_COMMANDS"))
)
config_dict["shell_allowlist"] = _safe_split(
os.getenv("SHELL_ALLOWLIST", os.getenv("ALLOW_COMMANDS"))
)
config_dict["google_custom_search_engine_id"] = os.getenv(
"GOOGLE_CUSTOM_SEARCH_ENGINE_ID", os.getenv("CUSTOM_SEARCH_ENGINE_ID")
)
config_dict["elevenlabs_voice_id"] = os.getenv(
"ELEVENLABS_VOICE_ID", os.getenv("ELEVENLABS_VOICE_1_ID")
)
if not config_dict["text_to_speech_provider"]:
if os.getenv("USE_MAC_OS_TTS"):
default_tts_provider = "macos"
elif config_dict["elevenlabs_api_key"]:
default_tts_provider = "elevenlabs"
elif os.getenv("USE_BRIAN_TTS"):
default_tts_provider = "streamelements"
else:
default_tts_provider = "gtts"
config_dict["text_to_speech_provider"] = default_tts_provider
config_dict["plugins_allowlist"] = _safe_split(os.getenv("ALLOWLISTED_PLUGINS"))
config_dict["plugins_denylist"] = _safe_split(os.getenv("DENYLISTED_PLUGINS"))
with contextlib.suppress(TypeError):
config_dict["image_size"] = int(os.getenv("IMAGE_SIZE"))
with contextlib.suppress(TypeError):
config_dict["redis_port"] = int(os.getenv("REDIS_PORT"))
with contextlib.suppress(TypeError):
config_dict["temperature"] = float(os.getenv("TEMPERATURE"))
if config_dict["use_azure"]:
azure_config = cls.load_azure_config(
workdir / config_dict["azure_config_file"]
)
config_dict.update(azure_config)
elif os.getenv("OPENAI_API_BASE_URL"):
config_dict["openai_api_base"] = os.getenv("OPENAI_API_BASE_URL")
openai_organization = os.getenv("OPENAI_ORGANIZATION")
if openai_organization is not None:
config_dict["openai_organization"] = openai_organization
config_dict_without_none_values = {
k: v for k, v in config_dict.items() if v is not None
}
config = cls.build_agent_configuration(config_dict_without_none_values)
# Set secondary config variables (that depend on other config variables)
config.plugins_config = PluginsConfig.load_config(
config.workdir / config.plugins_config_file,
config.plugins_denylist,
config.plugins_allowlist,
)
return config
@classmethod
def load_azure_config(cls, config_file: Path) -> Dict[str, str]:
"""
Loads the configuration parameters for Azure hosting from the specified file
path as a yaml file.
Parameters:
config_file (Path): The path to the config yaml file.
Returns:
Dict
"""
with open(config_file) as file:
config_params = yaml.load(file, Loader=yaml.FullLoader) or {}
return {
"openai_api_type": config_params.get("azure_api_type", "azure"),
"openai_api_base": config_params.get("azure_api_base", ""),
"openai_api_version": config_params.get(
"azure_api_version", "2023-03-15-preview"
),
"azure_model_to_deployment_id_map": config_params.get(
"azure_model_map", {}
),
}
def check_openai_api_key(config: Config) -> None:
"""Check if the OpenAI API key is set in config.py or as an environment variable."""
if not config.openai_api_key:
print(
Fore.RED
+ "Please set your OpenAI API key in .env or as an environment variable."
+ Fore.RESET
)
print("You can get your key from https://platform.openai.com/account/api-keys")
openai_api_key = input(
"If you do have the key, please enter your OpenAI API key now:\n"
)
key_pattern = r"^sk-\w{48}"
openai_api_key = openai_api_key.strip()
if re.search(key_pattern, openai_api_key):
os.environ["OPENAI_API_KEY"] = openai_api_key
config.openai_api_key = openai_api_key
print(
Fore.GREEN
+ "OpenAI API key successfully set!\n"
+ Fore.YELLOW
+ "NOTE: The API key you've set is only temporary.\n"
+ "For longer sessions, please set it in .env file"
+ Fore.RESET
)
else:
print("Invalid OpenAI API key!")
exit(1)
def _safe_split(s: Union[str, None], sep: str = ",") -> list[str]:
"""Split a string by a separator. Return an empty list if the string is None."""
if s is None:
return []
return s.split(sep)

View File

@@ -0,0 +1,47 @@
# sourcery skip: do-not-use-staticmethod
"""
A module that contains the PromptConfig class object that contains the configuration
"""
import yaml
from colorama import Fore
from autogpt import utils
from autogpt.logs import logger
class PromptConfig:
"""
A class object that contains the configuration information for the prompt, which will be used by the prompt generator
Attributes:
constraints (list): Constraints list for the prompt generator.
resources (list): Resources list for the prompt generator.
performance_evaluations (list): Performance evaluation list for the prompt generator.
"""
def __init__(self, prompt_settings_file: str) -> None:
"""
Initialize a class instance with parameters (constraints, resources, performance_evaluations) loaded from
yaml file if yaml file exists,
else raises error.
Parameters:
constraints (list): Constraints list for the prompt generator.
resources (list): Resources list for the prompt generator.
performance_evaluations (list): Performance evaluation list for the prompt generator.
Returns:
None
"""
# Validate file
(validated, message) = utils.validate_yaml_file(prompt_settings_file)
if not validated:
logger.typewriter_log("FAILED FILE VALIDATION", Fore.RED, message)
logger.double_check()
exit(1)
with open(prompt_settings_file, encoding="utf-8") as file:
config_params = yaml.load(file, Loader=yaml.FullLoader)
self.constraints = config_params.get("constraints", [])
self.resources = config_params.get("resources", [])
self.performance_evaluations = config_params.get("performance_evaluations", [])

View File

@@ -0,0 +1,272 @@
# Re-architecture Notes
## Key Documents
- [Planned Agent Workflow](https://whimsical.com/agent-workflow-v2-NmnTQ8R7sVo7M3S43XgXmZ)
- [Original Architecture Diagram](https://www.figma.com/file/fwdj44tPR7ArYtnGGUKknw/Modular-Architecture?type=whiteboard&node-id=0-1) - This is sadly well out of date at this point.
- [Kanban](https://github.com/orgs/Significant-Gravitas/projects/1/views/1?filterQuery=label%3Are-arch)
## The Motivation
The `master` branch of Auto-GPT is an organically grown amalgamation of many thoughts
and ideas about agent-driven autonomous systems. It lacks clear abstraction boundaries,
has issues of global state and poorly encapsulated state, and is generally just hard to
make effective changes to. Mainly it's just a system that's hard to make changes to.
And research in the field is moving fast, so we want to be able to try new ideas
quickly.
## Initial Planning
A large group of maintainers and contributors met do discuss the architectural
challenges associated with the existing codebase. Many much-desired features (building
new user interfaces, enabling project-specific agents, enabling multi-agent systems)
are bottlenecked by the global state in the system. We discussed the tradeoffs between
an incremental system transition and a big breaking version change and decided to go
for the breaking version change. We justified this by saying:
- We can maintain, in essence, the same user experience as now even with a radical
restructuring of the codebase
- Our developer audience is struggling to use the existing codebase to build
applications and libraries of their own, so this breaking change will largely be
welcome.
## Primary Goals
- Separate the AutoGPT application code from the library code.
- Remove global state from the system
- Allow for multiple agents per user (with facilities for running simultaneously)
- Create a serializable representation of an Agent
- Encapsulate the core systems in abstractions with clear boundaries.
## Secondary goals
- Use existing tools to ditch any unneccesary cruft in the codebase (document loading,
json parsing, anything easier to replace than to port).
- Bring in the [core agent loop updates](https://whimsical.com/agent-workflow-v2-NmnTQ8R7sVo7M3S43XgXmZ)
being developed simultaneously by @Pwuts
# The Agent Subsystems
## Configuration
We want a lot of things from a configuration system. We lean heavily on it in the
`master` branch to allow several parts of the system to communicate with each other.
[Recent work](https://github.com/Significant-Gravitas/Auto-GPT/pull/4737) has made it
so that the config is no longer a singleton object that is materialized from the import
state, but it's still treated as a
[god object](https://en.wikipedia.org/wiki/God_object) containing all information about
the system and _critically_ allowing any system to reference configuration information
about other parts of the system.
### What we want
- It should still be reasonable to collate the entire system configuration in a
sensible way.
- The configuration should be validatable and validated.
- The system configuration should be a _serializable_ representation of an `Agent`.
- The configuration system should provide a clear (albeit very low-level) contract
about user-configurable aspects of the system.
- The configuration should reasonably manage default values and user-provided overrides.
- The configuration system needs to handle credentials in a reasonable way.
- The configuration should be the representation of some amount of system state, like
api budgets and resource usage. These aspects are recorded in the configuration and
updated by the system itself.
- Agent systems should have encapsulated views of the configuration. E.g. the memory
system should know about memory configuration but nothing about command configuration.
## Workspace
There are two ways to think about the workspace:
- The workspace is a scratch space for an agent where it can store files, write code,
and do pretty much whatever else it likes.
- The workspace is, at any given point in time, the single source of truth for what an
agent is. It contains the serializable state (the configuration) as well as all
other working state (stored files, databases, memories, custom code).
In the existing system there is **one** workspace. And because the workspace holds so
much agent state, that means a user can only work with one agent at a time.
## Memory
The memory system has been under extremely active development.
See [#3536](https://github.com/Significant-Gravitas/Auto-GPT/issues/3536) and
[#4208](https://github.com/Significant-Gravitas/Auto-GPT/pull/4208) for discussion and
work in the `master` branch. The TL;DR is
that we noticed a couple of months ago that the `Agent` performed **worse** with
permanent memory than without it. Since then the knowledge storage and retrieval
system has been [redesigned](https://whimsical.com/memory-system-8Ae6x6QkjDwQAUe9eVJ6w1)
and partially implemented in the `master` branch.
## Planning/Prompt-Engineering
The planning system is the system that translates user desires/agent intentions into
language model prompts. In the course of development, it has become pretty clear
that `Planning` is the wrong name for this system
### What we want
- It should be incredibly obvious what's being passed to a language model, when it's
being passed, and what the language model response is. The landscape of language
model research is developing very rapidly, so building complex abstractions between
users/contributors and the language model interactions is going to make it very
difficult for us to nimbly respond to new research developments.
- Prompt-engineering should ideally be exposed in a parameterizeable way to users.
- We should, where possible, leverage OpenAI's new
[function calling api](https://openai.com/blog/function-calling-and-other-api-updates)
to get outputs in a standard machine-readable format and avoid the deep pit of
parsing json (and fixing unparsable json).
### Planning Strategies
The [new agent workflow](https://whimsical.com/agent-workflow-v2-NmnTQ8R7sVo7M3S43XgXmZ)
has many, many interaction points for language models. We really would like to not
distribute prompt templates and raw strings all through the system. The re-arch solution
is to encapsulate language model interactions into planning strategies.
These strategies are defined by
- The `LanguageModelClassification` they use (`FAST` or `SMART`)
- A function `build_prompt` that takes strategy specific arguments and constructs a
`LanguageModelPrompt` (a simple container for lists of messages and functions to
pass to the language model)
- A function `parse_content` that parses the response content (a dict) into a better
formatted dict. Contracts here are intentionally loose and will tighten once we have
at least one other language model provider.
## Resources
Resources are kinds of services we consume from external APIs. They may have associated
credentials and costs we need to manage. Management of those credentials is implemented
as manipulation of the resource configuration. We have two categories of resources
currently
- AI/ML model providers (including language model providers and embedding model providers, ie OpenAI)
- Memory providers (e.g. Pinecone, Weaviate, ChromaDB, etc.)
### What we want
- Resource abstractions should provide a common interface to different service providers
for a particular kind of service.
- Resource abstractions should manipulate the configuration to manage their credentials
and budget/accounting.
- Resource abstractions should be composable over an API (e.g. I should be able to make
an OpenAI provider that is both a LanguageModelProvider and an EmbeddingModelProvider
and use it wherever I need those services).
## Abilities
Along with planning and memory usage, abilities are one of the major augmentations of
augmented language models. They allow us to expand the scope of what language models
can do by hooking them up to code they can execute to obtain new knowledge or influence
the world.
### What we want
- Abilities should have an extremely clear interface that users can write to.
- Abilities should have an extremely clear interface that a language model can
understand
- Abilities should be declarative about their dependencies so the system can inject them
- Abilities should be executable (where sensible) in an async run loop.
- Abilities should be not have side effects unless those side effects are clear in
their representation to an agent (e.g. the BrowseWeb ability shouldn't write a file,
but the WriteFile ability can).
## Plugins
Users want to add lots of features that we don't want to support as first-party.
Or solution to this is a plugin system to allow users to plug in their functionality or
to construct their agent from a public plugin marketplace. Our primary concern in the
re-arch is to build a stateless plugin service interface and a simple implementation
that can load plugins from installed packages or from zip files. Future efforts will
expand this system to allow plugins to load from a marketplace or some other kind
of service.
### What is a Plugin
Plugins are a kind of garbage term. They refer to a number of things.
- New commands for the agent to execute. This is the most common usage.
- Replacements for entire subsystems like memory or language model providers
- Application plugins that do things like send emails or communicate via whatsapp
- The repositories contributors create that may themselves have multiple plugins in them.
### Usage in the existing system
The current plugin system is _hook-based_. This means plugins don't correspond to
kinds of objects in the system, but rather to times in the system at which we defer
execution to them. The main advantage of this setup is that user code can hijack
pretty much any behavior of the agent by injecting code that supercedes the normal
agent execution. The disadvantages to this approach are numerous:
- We have absolutely no mechanisms to enforce any security measures because the threat
surface is everything.
- We cannot reason about agent behavior in a cohesive way because control flow can be
ceded to user code at pretty much any point and arbitrarily change or break the
agent behavior
- The interface for designing a plugin is kind of terrible and difficult to standardize
- The hook based implementation means we couple ourselves to a particular flow of
control (or otherwise risk breaking plugin behavior). E.g. many of the hook targets
in the [old workflow](https://whimsical.com/agent-workflow-VAzeKcup3SR7awpNZJKTyK)
are not present or mean something entirely different in the
[new workflow](https://whimsical.com/agent-workflow-v2-NmnTQ8R7sVo7M3S43XgXmZ).
- Etc.
### What we want
- A concrete definition of a plugin that is narrow enough in scope that we can define
it well and reason about how it will work in the system.
- A set of abstractions that let us define a plugin by its storage format and location
- A service interface that knows how to parse the plugin abstractions and turn them
into concrete classes and objects.
## Some Notes on how and why we'll use OO in this project
First and foremost, Python itself is an object-oriented language. It's
underlying [data model](https://docs.python.org/3/reference/datamodel.html) is built
with object-oriented programming in mind. It offers useful tools like abstract base
classes to communicate interfaces to developers who want to, e.g., write plugins, or
help work on implementations. If we were working in a different language that offered
different tools, we'd use a different paradigm.
While many things are classes in the re-arch, they are not classes in the same way.
There are three kinds of things (roughly) that are written as classes in the re-arch:
1. **Configuration**: Auto-GPT has *a lot* of configuration. This configuration
is *data* and we use **[Pydantic](https://docs.pydantic.dev/latest/)** to manage it as
pydantic is basically industry standard for this stuff. It provides runtime validation
for all the configuration and allows us to easily serialize configuration to both basic
python types (dicts, lists, and primatives) as well as serialize to json, which is
important for us being able to put representations of agents
[on the wire](https://en.wikipedia.org/wiki/Wire_protocol) for web applications and
agent-to-agent communication. *These are essentially
[structs](https://en.wikipedia.org/wiki/Struct_(C_programming_language)) rather than
traditional classes.*
2. **Internal Data**: Very similar to configuration, Auto-GPT passes around boatloads
of internal data. We are interacting with language models and language model APIs
which means we are handling lots of *structured* but *raw* text. Here we also
leverage **pydantic** to both *parse* and *validate* the internal data and also to
give us concrete types which we can use static type checkers to validate against
and discover problems before they show up as bugs at runtime. *These are
essentially [structs](https://en.wikipedia.org/wiki/Struct_(C_programming_language))
rather than traditional classes.*
3. **System Interfaces**: This is our primary traditional use of classes in the
re-arch. We have a bunch of systems. We want many of those systems to have
alternative implementations (e.g. via plugins). We use abstract base classes to
define interfaces to communicate with people who might want to provide those
plugins. We provide a single concrete implementation of most of those systems as a
subclass of the interface. This should not be controversial.
The approach is consistent with
[prior](https://github.com/Significant-Gravitas/Auto-GPT/issues/2458)
[work](https://github.com/Significant-Gravitas/Auto-GPT/pull/2442) done by other
maintainers in this direction.
From an organization standpoint, OO programming is by far the most popular programming
paradigm (especially for Python). It's the one most often taught in programming classes
and the one with the most available online training for people interested in
contributing.
Finally, and importantly, we scoped the plan and initial design of the re-arch as a
large group of maintainers and collaborators early on. This is consistent with the
design we chose and no-one offered alternatives.

94
autogpt/core/README.md Normal file
View File

@@ -0,0 +1,94 @@
# Auto-GPT Core
This subpackage contains the ongoing work for the
[Auto-GPT Re-arch](https://github.com/Significant-Gravitas/Auto-GPT/issues/4770). It is
a work in progress and is not yet feature complete. In particular, it does not yet
have many of the Auto-GPT commands implemented and is pending ongoing work to
[re-incorporate vector-based memory and knowledge retrieval](https://github.com/Significant-Gravitas/Auto-GPT/issues/3536).
## [Overview](ARCHITECTURE_NOTES.md)
The Auto-GPT Re-arch is a re-implementation of the Auto-GPT agent that is designed to be more modular,
more extensible, and more maintainable than the original Auto-GPT agent. It is also designed to be
more accessible to new developers and to be easier to contribute to. The re-arch is a work in progress
and is not yet feature complete. It is also not yet ready for production use.
## Running the Re-arch Code
There are two client applications for Auto-GPT included.
Unlike the main version of Auto-GPT, the re-arch requires you to actually install Auto-GPT in your python
environment to run this application. To do so, run
```
pip install -e REPOSITORY_ROOT
```
where `REPOSITORY_ROOT` is the root of the Auto-GPT repository on your machine. The `REPOSITORY_ROOT`
is the directory that contains the `setup.py` file and is the main, top-level directory of the repository
when you clone it.
## CLI Application
:star2: **This is the reference application I'm working with for now** :star2:
The first app is a straight CLI application. I have not done anything yet to port all the friendly display stuff from the `logger.typewriter_log` logic.
- [Entry Point](https://github.com/Significant-Gravitas/Auto-GPT/blob/master/autogpt/core/runner/cli_app/cli.py)
- [Client Application](https://github.com/Significant-Gravitas/Auto-GPT/blob/master/autogpt/core/runner/cli_app/main.py)
You'll then need a settings file. Run
```
python REPOSITORY_ROOT/autogpt/core/runner/cli_app/cli.py make-settings
```
This will write a file called `default_agent_settings.yaml` with all the user-modifiable
configuration keys to `~/auto-gpt/default_agent_settings.yml` and make the `auto-gpt` directory
in your user directory if it doesn't exist). Your user directory is located in different places
depending on your operating system:
- On Linux, it's `/home/USERNAME`
- On Windows, it's `C:\Users\USERNAME`
- On Mac, it's `/Users/USERNAME`
At a bare minimum, you'll need to set `openai.credentials.api_key` to your OpenAI API Key to run
the model.
You can then run Auto-GPT with
```
python REPOSITORY_ROOT/autogpt/core/runner/cli_app/cli.py run
```
to launch the interaction loop.
### CLI Web App
:warning: I am not actively developing this application. I am primarily working with the traditional CLI app
described above. It is a very good place to get involved if you have web application design experience and are
looking to get involved in the re-arch.
The second app is still a CLI, but it sets up a local webserver that the client application talks to
rather than invoking calls to the Agent library code directly. This application is essentially a sketch
at this point as the folks who were driving it have had less time (and likely not enough clarity) to proceed.
- [Entry Point](https://github.com/Significant-Gravitas/Auto-GPT/blob/master/autogpt/core/runner/cli_web_app/cli.py)
- [Client Application](https://github.com/Significant-Gravitas/Auto-GPT/blob/master/autogpt/core/runner/cli_web_app/client/client.py)
- [Server API](https://github.com/Significant-Gravitas/Auto-GPT/blob/master/autogpt/core/runner/cli_web_app/server/api.py)
To run, you still need to generate a default configuration. You can do
```
python REPOSITORY_ROOT/autogpt/core/runner/cli_web_app/cli.py make-settings
```
It invokes the same command as the bare CLI app, so follow the instructions above about setting your API key.
To run, do
```
python REPOSITORY_ROOT/autogpt/core/runner/cli_web_app/cli.py client
```
This will launch a webserver and then start the client cli application to communicate with it.

View File

@@ -0,0 +1,4 @@
"""The command system provides a way to extend the functionality of the AI agent."""
from autogpt.core.ability.base import Ability, AbilityRegistry
from autogpt.core.ability.schema import AbilityResult
from autogpt.core.ability.simple import AbilityRegistrySettings, SimpleAbilityRegistry

View File

@@ -0,0 +1,92 @@
import abc
from pprint import pformat
from typing import ClassVar
import inflection
from pydantic import Field
from autogpt.core.ability.schema import AbilityResult
from autogpt.core.configuration import SystemConfiguration
from autogpt.core.planning.simple import LanguageModelConfiguration
class AbilityConfiguration(SystemConfiguration):
"""Struct for model configuration."""
from autogpt.core.plugin.base import PluginLocation
location: PluginLocation
packages_required: list[str] = Field(default_factory=list)
language_model_required: LanguageModelConfiguration = None
memory_provider_required: bool = False
workspace_required: bool = False
class Ability(abc.ABC):
"""A class representing an agent ability."""
default_configuration: ClassVar[AbilityConfiguration]
@classmethod
def name(cls) -> str:
"""The name of the ability."""
return inflection.underscore(cls.__name__)
@classmethod
@abc.abstractmethod
def description(cls) -> str:
"""A detailed description of what the ability does."""
...
@classmethod
@abc.abstractmethod
def arguments(cls) -> dict:
"""A dict of arguments in standard json schema format."""
...
@classmethod
def required_arguments(cls) -> list[str]:
"""A list of required arguments."""
return []
@abc.abstractmethod
async def __call__(self, *args, **kwargs) -> AbilityResult:
...
def __str__(self) -> str:
return pformat(self.dump)
def dump(self) -> dict:
return {
"name": self.name(),
"description": self.description(),
"parameters": {
"type": "object",
"properties": self.arguments(),
"required": self.required_arguments(),
},
}
class AbilityRegistry(abc.ABC):
@abc.abstractmethod
def register_ability(
self, ability_name: str, ability_configuration: AbilityConfiguration
) -> None:
...
@abc.abstractmethod
def list_abilities(self) -> list[str]:
...
@abc.abstractmethod
def dump_abilities(self) -> list[dict]:
...
@abc.abstractmethod
def get_ability(self, ability_name: str) -> Ability:
...
@abc.abstractmethod
def perform(self, ability_name: str, **kwargs) -> AbilityResult:
...

View File

@@ -0,0 +1,6 @@
from autogpt.core.ability.builtins.create_new_ability import CreateNewAbility
from autogpt.core.ability.builtins.query_language_model import QueryLanguageModel
BUILTIN_ABILITIES = {
QueryLanguageModel.name(): QueryLanguageModel,
}

View File

@@ -0,0 +1,102 @@
import logging
from autogpt.core.ability.base import Ability, AbilityConfiguration
from autogpt.core.ability.schema import AbilityResult
from autogpt.core.plugin.simple import PluginLocation, PluginStorageFormat
class CreateNewAbility(Ability):
default_configuration = AbilityConfiguration(
location=PluginLocation(
storage_format=PluginStorageFormat.INSTALLED_PACKAGE,
storage_route="autogpt.core.ability.builtins.CreateNewAbility",
),
)
def __init__(
self,
logger: logging.Logger,
configuration: AbilityConfiguration,
):
self._logger = logger
self._configuration = configuration
@classmethod
def description(cls) -> str:
return "Create a new ability by writing python code."
@classmethod
def arguments(cls) -> dict:
return {
"ability_name": {
"type": "string",
"description": "A meaningful and concise name for the new ability.",
},
"description": {
"type": "string",
"description": "A detailed description of the ability and its uses, including any limitations.",
},
"arguments": {
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {
"type": "string",
"description": "The name of the argument.",
},
"type": {
"type": "string",
"description": "The type of the argument. Must be a standard json schema type.",
},
"description": {
"type": "string",
"description": "A detailed description of the argument and its uses.",
},
},
},
"description": "A list of arguments that the ability will accept.",
},
"required_arguments": {
"type": "array",
"items": {
"type": "string",
"description": "The names of the arguments that are required.",
},
"description": "A list of the names of the arguments that are required.",
},
"package_requirements": {
"type": "array",
"items": {
"type": "string",
"description": "The of the Python package that is required to execute the ability.",
},
"description": "A list of the names of the Python packages that are required to execute the ability.",
},
"code": {
"type": "string",
"description": "The Python code that will be executed when the ability is called.",
},
}
@classmethod
def required_arguments(cls) -> list[str]:
return [
"ability_name",
"description",
"arguments",
"required_arguments",
"package_requirements",
"code",
]
async def __call__(
self,
ability_name: str,
description: str,
arguments: list[dict],
required_arguments: list[str],
package_requirements: list[str],
code: str,
) -> AbilityResult:
raise NotImplementedError

View File

@@ -0,0 +1,167 @@
import logging
import os
from autogpt.core.ability.base import Ability, AbilityConfiguration
from autogpt.core.ability.schema import AbilityResult, ContentType, Knowledge
from autogpt.core.workspace import Workspace
class ReadFile(Ability):
default_configuration = AbilityConfiguration(
packages_required=["unstructured"],
workspace_required=True,
)
def __init__(
self,
logger: logging.Logger,
workspace: Workspace,
):
self._logger = logger
self._workspace = workspace
@property
def description(self) -> str:
return "Read and parse all text from a file."
@property
def arguments(self) -> dict:
return {
"filename": {
"type": "string",
"description": "The name of the file to read.",
},
}
def _check_preconditions(self, filename: str) -> AbilityResult | None:
message = ""
try:
pass
except ImportError:
message = "Package charset_normalizer is not installed."
try:
file_path = self._workspace.get_path(filename)
if not file_path.exists():
message = f"File {filename} does not exist."
if not file_path.is_file():
message = f"{filename} is not a file."
except ValueError as e:
message = str(e)
if message:
return AbilityResult(
ability_name=self.name(),
ability_args={"filename": filename},
success=False,
message=message,
data=None,
)
def __call__(self, filename: str) -> AbilityResult:
if result := self._check_preconditions(filename):
return result
from unstructured.partition.auto import partition
file_path = self._workspace.get_path(filename)
try:
elements = partition(str(file_path))
# TODO: Lots of other potentially useful information is available
# in the partitioned file. Consider returning more of it.
new_knowledge = Knowledge(
content="\n\n".join([element.text for element in elements]),
content_type=ContentType.TEXT,
content_metadata={"filename": filename},
)
success = True
message = f"File {file_path} read successfully."
except IOError as e:
new_knowledge = None
success = False
message = str(e)
return AbilityResult(
ability_name=self.name(),
ability_args={"filename": filename},
success=success,
message=message,
new_knowledge=new_knowledge,
)
class WriteFile(Ability):
default_configuration = AbilityConfiguration(
packages_required=["unstructured"],
workspace_required=True,
)
def __init__(
self,
logger: logging.Logger,
workspace: Workspace,
):
self._logger = logger
self._workspace = workspace
@property
def description(self) -> str:
return "Write text to a file."
@property
def arguments(self) -> dict:
return {
"filename": {
"type": "string",
"description": "The name of the file to write.",
},
"contents": {
"type": "string",
"description": "The contents of the file to write.",
},
}
def _check_preconditions(
self, filename: str, contents: str
) -> AbilityResult | None:
message = ""
try:
file_path = self._workspace.get_path(filename)
if file_path.exists():
message = f"File {filename} already exists."
if len(contents):
message = f"File {filename} was not given any content."
except ValueError as e:
message = str(e)
if message:
return AbilityResult(
ability_name=self.name(),
ability_args={"filename": filename, "contents": contents},
success=False,
message=message,
data=None,
)
def __call__(self, filename: str, contents: str) -> AbilityResult:
if result := self._check_preconditions(filename, contents):
return result
file_path = self._workspace.get_path(filename)
try:
directory = os.path.dirname(file_path)
os.makedirs(directory)
with open(filename, "w", encoding="utf-8") as f:
f.write(contents)
success = True
message = f"File {file_path} written successfully."
except IOError as e:
success = False
message = str(e)
return AbilityResult(
ability_name=self.name(),
ability_args={"filename": filename},
success=success,
message=message,
)

View File

@@ -0,0 +1,78 @@
import logging
from autogpt.core.ability.base import Ability, AbilityConfiguration
from autogpt.core.ability.schema import AbilityResult
from autogpt.core.planning.simple import LanguageModelConfiguration
from autogpt.core.plugin.simple import PluginLocation, PluginStorageFormat
from autogpt.core.resource.model_providers import (
LanguageModelMessage,
LanguageModelProvider,
MessageRole,
ModelProviderName,
OpenAIModelName,
)
class QueryLanguageModel(Ability):
default_configuration = AbilityConfiguration(
location=PluginLocation(
storage_format=PluginStorageFormat.INSTALLED_PACKAGE,
storage_route="autogpt.core.ability.builtins.QueryLanguageModel",
),
language_model_required=LanguageModelConfiguration(
model_name=OpenAIModelName.GPT3,
provider_name=ModelProviderName.OPENAI,
temperature=0.9,
),
)
def __init__(
self,
logger: logging.Logger,
configuration: AbilityConfiguration,
language_model_provider: LanguageModelProvider,
):
self._logger = logger
self._configuration = configuration
self._language_model_provider = language_model_provider
@classmethod
def description(cls) -> str:
return "Query a language model. A query should be a question and any relevant context."
@classmethod
def arguments(cls) -> dict:
return {
"query": {
"type": "string",
"description": "A query for a language model. A query should contain a question and any relevant context.",
},
}
@classmethod
def required_arguments(cls) -> list[str]:
return ["query"]
async def __call__(self, query: str) -> AbilityResult:
messages = [
LanguageModelMessage(
content=query,
role=MessageRole.USER,
),
]
model_response = await self._language_model_provider.create_language_completion(
model_prompt=messages,
functions=[],
model_name=self._configuration.language_model_required.model_name,
completion_parser=self._parse_response,
)
return AbilityResult(
ability_name=self.name(),
ability_args={"query": query},
success=True,
message=model_response.content["content"],
)
@staticmethod
def _parse_response(response_content: dict) -> dict:
return {"content": response_content["content"]}

View File

@@ -0,0 +1,30 @@
import enum
from typing import Any
from pydantic import BaseModel
class ContentType(str, enum.Enum):
# TBD what these actually are.
TEXT = "text"
CODE = "code"
class Knowledge(BaseModel):
content: str
content_type: ContentType
content_metadata: dict[str, Any]
class AbilityResult(BaseModel):
"""The AbilityResult is a standard response struct for an ability."""
ability_name: str
ability_args: dict[str, str]
success: bool
message: str
new_knowledge: Knowledge = None
def summary(self) -> str:
kwargs = ", ".join(f"{k}={v}" for k, v in self.ability_args.items())
return f"{self.ability_name}({kwargs}): {self.message}"

View File

@@ -0,0 +1,96 @@
import logging
from autogpt.core.ability.base import Ability, AbilityConfiguration, AbilityRegistry
from autogpt.core.ability.builtins import BUILTIN_ABILITIES
from autogpt.core.ability.schema import AbilityResult
from autogpt.core.configuration import Configurable, SystemConfiguration, SystemSettings
from autogpt.core.memory.base import Memory
from autogpt.core.plugin.simple import SimplePluginService
from autogpt.core.resource.model_providers import (
LanguageModelProvider,
ModelProviderName,
)
from autogpt.core.workspace.base import Workspace
class AbilityRegistryConfiguration(SystemConfiguration):
"""Configuration for the AbilityRegistry subsystem."""
abilities: dict[str, AbilityConfiguration]
class AbilityRegistrySettings(SystemSettings):
configuration: AbilityRegistryConfiguration
class SimpleAbilityRegistry(AbilityRegistry, Configurable):
default_settings = AbilityRegistrySettings(
name="simple_ability_registry",
description="A simple ability registry.",
configuration=AbilityRegistryConfiguration(
abilities={
ability_name: ability.default_configuration
for ability_name, ability in BUILTIN_ABILITIES.items()
},
),
)
def __init__(
self,
settings: AbilityRegistrySettings,
logger: logging.Logger,
memory: Memory,
workspace: Workspace,
model_providers: dict[ModelProviderName, LanguageModelProvider],
):
self._configuration = settings.configuration
self._logger = logger
self._memory = memory
self._workspace = workspace
self._model_providers = model_providers
self._abilities = []
for (
ability_name,
ability_configuration,
) in self._configuration.abilities.items():
self.register_ability(ability_name, ability_configuration)
def register_ability(
self, ability_name: str, ability_configuration: AbilityConfiguration
) -> None:
ability_class = SimplePluginService.get_plugin(ability_configuration.location)
ability_args = {
"logger": self._logger.getChild(ability_name),
"configuration": ability_configuration,
}
if ability_configuration.packages_required:
# TODO: Check packages are installed and maybe install them.
pass
if ability_configuration.memory_provider_required:
ability_args["memory"] = self._memory
if ability_configuration.workspace_required:
ability_args["workspace"] = self._workspace
if ability_configuration.language_model_required:
ability_args["language_model_provider"] = self._model_providers[
ability_configuration.language_model_required.provider_name
]
ability = ability_class(**ability_args)
self._abilities.append(ability)
def list_abilities(self) -> list[str]:
return [
f"{ability.name()}: {ability.description()}" for ability in self._abilities
]
def dump_abilities(self) -> list[dict]:
return [ability.dump() for ability in self._abilities]
def get_ability(self, ability_name: str) -> Ability:
for ability in self._abilities:
if ability.name() == ability_name:
return ability
raise ValueError(f"Ability '{ability_name}' not found.")
async def perform(self, ability_name: str, **kwargs) -> AbilityResult:
ability = self.get_ability(ability_name)
return await ability(**kwargs)

View File

@@ -0,0 +1,3 @@
"""The Agent is an autonomouos entity guided by a LLM provider."""
from autogpt.core.agent.base import Agent
from autogpt.core.agent.simple import AgentSettings, SimpleAgent

View File

@@ -0,0 +1,26 @@
import abc
import logging
from pathlib import Path
class Agent(abc.ABC):
@abc.abstractmethod
def __init__(self, *args, **kwargs):
...
@classmethod
@abc.abstractmethod
def from_workspace(
cls,
workspace_path: Path,
logger: logging.Logger,
) -> "Agent":
...
@abc.abstractmethod
async def determine_next_ability(self, *args, **kwargs):
...
@abc.abstractmethod
def __repr__(self):
...

View File

@@ -0,0 +1,391 @@
import logging
from datetime import datetime
from pathlib import Path
from typing import Any
from pydantic import BaseModel
from autogpt.core.ability import (
AbilityRegistrySettings,
AbilityResult,
SimpleAbilityRegistry,
)
from autogpt.core.agent.base import Agent
from autogpt.core.configuration import Configurable, SystemConfiguration, SystemSettings
from autogpt.core.memory import MemorySettings, SimpleMemory
from autogpt.core.planning import PlannerSettings, SimplePlanner, Task, TaskStatus
from autogpt.core.plugin.simple import (
PluginLocation,
PluginStorageFormat,
SimplePluginService,
)
from autogpt.core.resource.model_providers import OpenAIProvider, OpenAISettings
from autogpt.core.workspace.simple import SimpleWorkspace, WorkspaceSettings
class AgentSystems(SystemConfiguration):
ability_registry: PluginLocation
memory: PluginLocation
openai_provider: PluginLocation
planning: PluginLocation
workspace: PluginLocation
class AgentConfiguration(SystemConfiguration):
cycle_count: int
max_task_cycle_count: int
creation_time: str
name: str
role: str
goals: list[str]
systems: AgentSystems
class AgentSystemSettings(SystemSettings):
configuration: AgentConfiguration
class AgentSettings(BaseModel):
agent: AgentSystemSettings
ability_registry: AbilityRegistrySettings
memory: MemorySettings
openai_provider: OpenAISettings
planning: PlannerSettings
workspace: WorkspaceSettings
def update_agent_name_and_goals(self, agent_goals: dict) -> None:
self.agent.configuration.name = agent_goals["agent_name"]
self.agent.configuration.role = agent_goals["agent_role"]
self.agent.configuration.goals = agent_goals["agent_goals"]
class SimpleAgent(Agent, Configurable):
default_settings = AgentSystemSettings(
name="simple_agent",
description="A simple agent.",
configuration=AgentConfiguration(
name="Entrepreneur-GPT",
role=(
"An AI designed to autonomously develop and run businesses with "
"the sole goal of increasing your net worth."
),
goals=[
"Increase net worth",
"Grow Twitter Account",
"Develop and manage multiple businesses autonomously",
],
cycle_count=0,
max_task_cycle_count=3,
creation_time="",
systems=AgentSystems(
ability_registry=PluginLocation(
storage_format=PluginStorageFormat.INSTALLED_PACKAGE,
storage_route="autogpt.core.ability.SimpleAbilityRegistry",
),
memory=PluginLocation(
storage_format=PluginStorageFormat.INSTALLED_PACKAGE,
storage_route="autogpt.core.memory.SimpleMemory",
),
openai_provider=PluginLocation(
storage_format=PluginStorageFormat.INSTALLED_PACKAGE,
storage_route="autogpt.core.resource.model_providers.OpenAIProvider",
),
planning=PluginLocation(
storage_format=PluginStorageFormat.INSTALLED_PACKAGE,
storage_route="autogpt.core.planning.SimplePlanner",
),
workspace=PluginLocation(
storage_format=PluginStorageFormat.INSTALLED_PACKAGE,
storage_route="autogpt.core.workspace.SimpleWorkspace",
),
),
),
)
def __init__(
self,
settings: AgentSystemSettings,
logger: logging.Logger,
ability_registry: SimpleAbilityRegistry,
memory: SimpleMemory,
openai_provider: OpenAIProvider,
planning: SimplePlanner,
workspace: SimpleWorkspace,
):
self._configuration = settings.configuration
self._logger = logger
self._ability_registry = ability_registry
self._memory = memory
# FIXME: Need some work to make this work as a dict of providers
# Getting the construction of the config to work is a bit tricky
self._openai_provider = openai_provider
self._planning = planning
self._workspace = workspace
self._task_queue = []
self._completed_tasks = []
self._current_task = None
self._next_ability = None
@classmethod
def from_workspace(
cls,
workspace_path: Path,
logger: logging.Logger,
) -> "SimpleAgent":
agent_settings = SimpleWorkspace.load_agent_settings(workspace_path)
agent_args = {}
agent_args["settings"] = agent_settings.agent
agent_args["logger"] = logger
agent_args["workspace"] = cls._get_system_instance(
"workspace",
agent_settings,
logger,
)
agent_args["openai_provider"] = cls._get_system_instance(
"openai_provider",
agent_settings,
logger,
)
agent_args["planning"] = cls._get_system_instance(
"planning",
agent_settings,
logger,
model_providers={"openai": agent_args["openai_provider"]},
)
agent_args["memory"] = cls._get_system_instance(
"memory",
agent_settings,
logger,
workspace=agent_args["workspace"],
)
agent_args["ability_registry"] = cls._get_system_instance(
"ability_registry",
agent_settings,
logger,
workspace=agent_args["workspace"],
memory=agent_args["memory"],
model_providers={"openai": agent_args["openai_provider"]},
)
return cls(**agent_args)
async def build_initial_plan(self) -> dict:
plan = await self._planning.make_initial_plan(
agent_name=self._configuration.name,
agent_role=self._configuration.role,
agent_goals=self._configuration.goals,
abilities=self._ability_registry.list_abilities(),
)
tasks = [Task.parse_obj(task) for task in plan.content["task_list"]]
# TODO: Should probably do a step to evaluate the quality of the generated tasks,
# and ensure that they have actionable ready and acceptance criteria
self._task_queue.extend(tasks)
self._task_queue.sort(key=lambda t: t.priority, reverse=True)
self._task_queue[-1].context.status = TaskStatus.READY
return plan.content
async def determine_next_ability(self, *args, **kwargs):
if not self._task_queue:
return {"response": "I don't have any tasks to work on right now."}
self._configuration.cycle_count += 1
task = self._task_queue.pop()
self._logger.info(f"Working on task: {task}")
task = await self._evaluate_task_and_add_context(task)
next_ability = await self._choose_next_ability(
task,
self._ability_registry.dump_abilities(),
)
self._current_task = task
self._next_ability = next_ability.content
return self._current_task, self._next_ability
async def execute_next_ability(self, user_input: str, *args, **kwargs):
if user_input == "y":
ability = self._ability_registry.get_ability(
self._next_ability["next_ability"]
)
ability_response = await ability(**self._next_ability["ability_arguments"])
await self._update_tasks_and_memory(ability_response)
if self._current_task.context.status == TaskStatus.DONE:
self._completed_tasks.append(self._current_task)
else:
self._task_queue.append(self._current_task)
self._current_task = None
self._next_ability = None
return ability_response.dict()
else:
raise NotImplementedError
async def _evaluate_task_and_add_context(self, task: Task) -> Task:
"""Evaluate the task and add context to it."""
if task.context.status == TaskStatus.IN_PROGRESS:
# Nothing to do here
return task
else:
self._logger.debug(f"Evaluating task {task} and adding relevant context.")
# TODO: Look up relevant memories (need working memory system)
# TODO: Evaluate whether there is enough information to start the task (language model call).
task.context.enough_info = True
task.context.status = TaskStatus.IN_PROGRESS
return task
async def _choose_next_ability(self, task: Task, ability_schema: list[dict]):
"""Choose the next ability to use for the task."""
self._logger.debug(f"Choosing next ability for task {task}.")
if task.context.cycle_count > self._configuration.max_task_cycle_count:
# Don't hit the LLM, just set the next action as "breakdown_task" with an appropriate reason
raise NotImplementedError
elif not task.context.enough_info:
# Don't ask the LLM, just set the next action as "breakdown_task" with an appropriate reason
raise NotImplementedError
else:
next_ability = await self._planning.determine_next_ability(
task, ability_schema
)
return next_ability
async def _update_tasks_and_memory(self, ability_result: AbilityResult):
self._current_task.context.cycle_count += 1
self._current_task.context.prior_actions.append(ability_result)
# TODO: Summarize new knowledge
# TODO: store knowledge and summaries in memory and in relevant tasks
# TODO: evaluate whether the task is complete
def __repr__(self):
return "SimpleAgent()"
################################################################
# Factory interface for agent bootstrapping and initialization #
################################################################
@classmethod
def build_user_configuration(cls) -> dict[str, Any]:
"""Build the user's configuration."""
configuration_dict = {
"agent": cls.get_user_config(),
}
system_locations = configuration_dict["agent"]["configuration"]["systems"]
for system_name, system_location in system_locations.items():
system_class = SimplePluginService.get_plugin(system_location)
configuration_dict[system_name] = system_class.get_user_config()
configuration_dict = _prune_empty_dicts(configuration_dict)
return configuration_dict
@classmethod
def compile_settings(
cls, logger: logging.Logger, user_configuration: dict
) -> AgentSettings:
"""Compile the user's configuration with the defaults."""
logger.debug("Processing agent system configuration.")
configuration_dict = {
"agent": cls.build_agent_configuration(
user_configuration.get("agent", {})
).dict(),
}
system_locations = configuration_dict["agent"]["configuration"]["systems"]
# Build up default configuration
for system_name, system_location in system_locations.items():
logger.debug(f"Compiling configuration for system {system_name}")
system_class = SimplePluginService.get_plugin(system_location)
configuration_dict[system_name] = system_class.build_agent_configuration(
user_configuration.get(system_name, {})
).dict()
return AgentSettings.parse_obj(configuration_dict)
@classmethod
async def determine_agent_name_and_goals(
cls,
user_objective: str,
agent_settings: AgentSettings,
logger: logging.Logger,
) -> dict:
logger.debug("Loading OpenAI provider.")
provider: OpenAIProvider = cls._get_system_instance(
"openai_provider",
agent_settings,
logger=logger,
)
logger.debug("Loading agent planner.")
agent_planner: SimplePlanner = cls._get_system_instance(
"planning",
agent_settings,
logger=logger,
model_providers={"openai": provider},
)
logger.debug("determining agent name and goals.")
model_response = await agent_planner.decide_name_and_goals(
user_objective,
)
return model_response.content
@classmethod
def provision_agent(
cls,
agent_settings: AgentSettings,
logger: logging.Logger,
):
agent_settings.agent.configuration.creation_time = datetime.now().strftime(
"%Y%m%d_%H%M%S"
)
workspace: SimpleWorkspace = cls._get_system_instance(
"workspace",
agent_settings,
logger=logger,
)
return workspace.setup_workspace(agent_settings, logger)
@classmethod
def _get_system_instance(
cls,
system_name: str,
agent_settings: AgentSettings,
logger: logging.Logger,
*args,
**kwargs,
):
system_locations = agent_settings.agent.configuration.systems.dict()
system_settings = getattr(agent_settings, system_name)
system_class = SimplePluginService.get_plugin(system_locations[system_name])
system_instance = system_class(
system_settings,
*args,
logger=logger.getChild(system_name),
**kwargs,
)
return system_instance
def _prune_empty_dicts(d: dict) -> dict:
"""
Prune branches from a nested dictionary if the branch only contains empty dictionaries at the leaves.
Args:
d: The dictionary to prune.
Returns:
The pruned dictionary.
"""
pruned = {}
for key, value in d.items():
if isinstance(value, dict):
pruned_value = _prune_empty_dicts(value)
if (
pruned_value
): # if the pruned dictionary is not empty, add it to the result
pruned[key] = pruned_value
else:
pruned[key] = value
return pruned

View File

@@ -0,0 +1,7 @@
"""The configuration encapsulates settings for all Agent subsystems."""
from autogpt.core.configuration.schema import (
Configurable,
SystemConfiguration,
SystemSettings,
UserConfigurable,
)

View File

@@ -0,0 +1,107 @@
import abc
import typing
from typing import Any, Generic, TypeVar
from pydantic import BaseModel, Field
def UserConfigurable(*args, **kwargs):
return Field(*args, **kwargs, user_configurable=True)
class SystemConfiguration(BaseModel):
def get_user_config(self) -> dict[str, Any]:
return _get_user_config_fields(self)
class Config:
extra = "forbid"
use_enum_values = True
class SystemSettings(BaseModel):
"""A base class for all system settings."""
name: str
description: str
class Config:
extra = "forbid"
use_enum_values = True
S = TypeVar("S", bound=SystemSettings)
class Configurable(abc.ABC, Generic[S]):
"""A base class for all configurable objects."""
prefix: str = ""
default_settings: typing.ClassVar[S]
@classmethod
def get_user_config(cls) -> dict[str, Any]:
return _get_user_config_fields(cls.default_settings)
@classmethod
def build_agent_configuration(cls, configuration: dict) -> S:
"""Process the configuration for this object."""
defaults = cls.default_settings.dict()
final_configuration = deep_update(defaults, configuration)
return cls.default_settings.__class__.parse_obj(final_configuration)
def _get_user_config_fields(instance: BaseModel) -> dict[str, Any]:
"""
Get the user config fields of a Pydantic model instance.
Args:
instance: The Pydantic model instance.
Returns:
The user config fields of the instance.
"""
user_config_fields = {}
for name, value in instance.__dict__.items():
field_info = instance.__fields__[name]
if "user_configurable" in field_info.field_info.extra:
user_config_fields[name] = value
elif isinstance(value, SystemConfiguration):
user_config_fields[name] = value.get_user_config()
elif isinstance(value, list) and all(
isinstance(i, SystemConfiguration) for i in value
):
user_config_fields[name] = [i.get_user_config() for i in value]
elif isinstance(value, dict) and all(
isinstance(i, SystemConfiguration) for i in value.values()
):
user_config_fields[name] = {
k: v.get_user_config() for k, v in value.items()
}
return user_config_fields
def deep_update(original_dict: dict, update_dict: dict) -> dict:
"""
Recursively update a dictionary.
Args:
original_dict (dict): The dictionary to be updated.
update_dict (dict): The dictionary to update with.
Returns:
dict: The updated dictionary.
"""
for key, value in update_dict.items():
if (
key in original_dict
and isinstance(original_dict[key], dict)
and isinstance(value, dict)
):
original_dict[key] = deep_update(original_dict[key], value)
else:
original_dict[key] = value
return original_dict

View File

@@ -0,0 +1,3 @@
"""The memory subsystem manages the Agent's long-term memory."""
from autogpt.core.memory.base import Memory
from autogpt.core.memory.simple import MemorySettings, SimpleMemory

View File

@@ -0,0 +1,13 @@
import abc
class Memory(abc.ABC):
pass
class MemoryItem(abc.ABC):
pass
class MessageHistory(abc.ABC):
pass

View File

@@ -0,0 +1,47 @@
import json
import logging
from autogpt.core.configuration import Configurable, SystemConfiguration, SystemSettings
from autogpt.core.memory.base import Memory
from autogpt.core.workspace import Workspace
class MemoryConfiguration(SystemConfiguration):
pass
class MemorySettings(SystemSettings):
configuration: MemoryConfiguration
class MessageHistory:
def __init__(self, previous_message_history: list[str]):
self._message_history = previous_message_history
class SimpleMemory(Memory, Configurable):
default_settings = MemorySettings(
name="simple_memory",
description="A simple memory.",
configuration=MemoryConfiguration(),
)
def __init__(
self,
settings: MemorySettings,
logger: logging.Logger,
workspace: Workspace,
):
self._configuration = settings.configuration
self._logger = logger
self._message_history = self._load_message_history(workspace)
@staticmethod
def _load_message_history(workspace: Workspace):
message_history_path = workspace.get_path("message_history.json")
if message_history_path.exists():
with message_history_path.open("r") as f:
message_history = json.load(f)
else:
message_history = []
return MessageHistory(message_history)

View File

@@ -0,0 +1,10 @@
"""The planning system organizes the Agent's activities."""
from autogpt.core.planning.schema import (
LanguageModelClassification,
LanguageModelPrompt,
LanguageModelResponse,
Task,
TaskStatus,
TaskType,
)
from autogpt.core.planning.simple import PlannerSettings, SimplePlanner

View File

@@ -0,0 +1,76 @@
import abc
from autogpt.core.configuration import SystemConfiguration
from autogpt.core.planning.schema import (
LanguageModelClassification,
LanguageModelPrompt,
)
# class Planner(abc.ABC):
# """Manages the agent's planning and goal-setting by constructing language model prompts."""
#
# @staticmethod
# @abc.abstractmethod
# async def decide_name_and_goals(
# user_objective: str,
# ) -> LanguageModelResponse:
# """Decide the name and goals of an Agent from a user-defined objective.
#
# Args:
# user_objective: The user-defined objective for the agent.
#
# Returns:
# The agent name and goals as a response from the language model.
#
# """
# ...
#
# @abc.abstractmethod
# async def plan(self, context: PlanningContext) -> LanguageModelResponse:
# """Plan the next ability for the Agent.
#
# Args:
# context: A context object containing information about the agent's
# progress, result, memories, and feedback.
#
#
# Returns:
# The next ability the agent should take along with thoughts and reasoning.
#
# """
# ...
#
# @abc.abstractmethod
# def reflect(
# self,
# context: ReflectionContext,
# ) -> LanguageModelResponse:
# """Reflect on a planned ability and provide self-criticism.
#
#
# Args:
# context: A context object containing information about the agent's
# reasoning, plan, thoughts, and criticism.
#
# Returns:
# Self-criticism about the agent's plan.
#
# """
# ...
class PromptStrategy(abc.ABC):
default_configuration: SystemConfiguration
@property
@abc.abstractmethod
def model_classification(self) -> LanguageModelClassification:
...
@abc.abstractmethod
def build_prompt(self, *_, **kwargs) -> LanguageModelPrompt:
...
@abc.abstractmethod
def parse_response_content(self, response_content: dict) -> dict:
...

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