mirror of
https://github.com/All-Hands-AI/OpenHands.git
synced 2026-04-29 03:00:45 -04:00
01ef90205d
* update swe_bench prompt; use minimal prompt for codeact; * upgrade agentskills and update testcases * update infer prompt * fix cwd * add icl for swebench * also log in_context_example to run infer * remove extra print * change prompt to abs path * update error message to include current file info * change cwd for jupyter if needed * update edit error message * update prompt * improve git get patch * update hint string * default to 50 turns * revert changes from codeact agent and create new CodeActSWEAgent * revert changes to codeact * revert instructions for run infer * revert instructions for run infer * update README * update max iter * add codeact swe agent * fix issue for CodeActSWEAgent * allow specifying max iter in cmdline script * stop printing * Update agenthub/codeact_swe_agent/README.md Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com> * Fix prompt regression in jupyter plugin --------- Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com> Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
CodeAct Agent Framework
This folder implements the CodeAct idea (paper, tweet) that consolidates LLM agents’ actions into a unified code action space for both simplicity and performance (see paper for more details).
The conceptual idea is illustrated below. At each turn, the agent can:
- Converse: Communicate with humans in natural language to ask for clarification, confirmation, etc.
- CodeAct: Choose to perform the task by executing code
- Execute any valid Linux
bashcommand - Execute any valid
Pythoncode with an interactive Python interpreter. This is simulated throughbashcommand, see plugin system below for more details.
- Execute any valid Linux
Plugin System
To make the CodeAct agent more powerful with only access to bash action space, CodeAct agent leverages OpenDevin's plugin system:
- Jupyter plugin: for IPython execution via bash command
- SWE-agent tool plugin: Powerful bash command line tools for software development tasks introduced by swe-agent.
Demo
https://github.com/OpenDevin/OpenDevin/assets/38853559/f592a192-e86c-4f48-ad31-d69282d5f6ac
Example of CodeActAgent with gpt-4-turbo-2024-04-09 performing a data science task (linear regression)
Work-in-progress & Next step
[] Support web-browsing [] Complete the workflow for CodeAct agent to submit Github PRs