xingyaoww/code-act
An LLM agent framework that unifies agent actions into an executable code action space integrated with a Python interpreter.
Not currently ranked — collecting fresh signals.
star history
The CodeAct approach consolidates LLM agent actions into a unified code-based action space, enabling agents to execute code, observe results, and dynamically revise or emit new actions through multi-turn interactions. The repository releases the CodeActAgent model (fine-tuned Mistral-7B), a multi-turn instruction dataset (CodeActInstruct), and a benchmark (M³ToolEval). It supports serving via Ollama, llama.cpp for local laptop inference, and Kubernetes deployment.
Frequently asked
- What is xingyaoww/code-act?
- An LLM agent framework that unifies agent actions into an executable code action space integrated with a Python interpreter.
- Is code-act open source?
- Yes — xingyaoww/code-act is open source, released under the MIT license.
- What language is code-act written in?
- xingyaoww/code-act is primarily written in Python.
- How popular is code-act?
- xingyaoww/code-act has 1.7k stars on GitHub.
- Where can I find code-act?
- xingyaoww/code-act is on GitHub at https://github.com/xingyaoww/code-act.