mirror of
https://github.com/OpenHands/OpenHands.git
synced 2025-12-26 05:48:36 +08:00
242c4a0df6
* remove extra actions * remove message observations * support null obs * handle null obs * fix frontend for changes * fix the way messages flow to the UI * change think to message * add regen script * regenerate all integration tests * change task * remove gh test * fix messages * fix tests * help agent exit after hitting max iter * Update opendevin/events/observation/success.py Co-authored-by: Engel Nyst <enyst@users.noreply.github.com> * Update agenthub/codeact_agent/codeact_agent.py Co-authored-by: Engel Nyst <enyst@users.noreply.github.com> --------- Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
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