Experimental API for running LLM agents (LangGraph, CrewAI, etc.) with distributed tool execution on Ray.
Note: This is an experimental package and currently in active development.
- Distributed tool execution on Ray clusters
- Framework-agnostic agent adapters (LangGraph, etc.)
- Code Interpreter: Secure Python execution in gVisor-sandboxed environments (see CODE_INTERPRETER.md)
- Session management with conversation history
macOS with Docker Desktop:
make setup-gvisor-macos
# Then quit and restart Docker Desktop
make verify-gvisorLinux:
sudo apt-get update && sudo apt-get install -y runsc
sudo runsc install
docker run --rm --runtime=runsc hello-worldThis project is licensed under the Apache License 2.0 - see the LICENSE file for details.
Contributions are welcome! Please feel free to submit a Pull Request.