OSS platform for AI software dev agents
Open-source sandboxed platform for running AI software engineering agents. Supports multiple agent frameworks and LLMs.
A fully autonomous agent you assign tasks to — it handles entire features or PRs
Other tools in this slot:
AIchitect's Genome scanner detects OpenHands in your project via these signals:
openhands-ai.openhands/OpenHands runs all agent tasks in E2B cloud sandboxes by default, isolating file writes and command execution from the host.
→ Safe, observable autonomous coding — every agent action happens in a reproducible cloud sandbox with no local environment risk.
OpenHands routes all LLM calls through LiteLLM, making it model-agnostic — swap Claude, GPT-4o, or a local model via config.
→ Model flexibility for autonomous agents without changing OpenHands code — route to the best model per task type.
OpenHands uses LangChain tool interfaces for its agent scaffolding, giving its agents access to LangChain's tool and retrieval ecosystem.
→ OpenHands agents can use any LangChain tool — vector retrieval, API calls, and data transforms — within autonomous task runs.
OpenHands connects to the GitHub MCP server to read repo state, commit code, and open PRs as part of autonomous development tasks.
→ End-to-end autonomous GitHub workflow: OpenHands reads issues, writes code, and opens a reviewed PR without human keyboard input.
Add to your GitHub README
[](https://aichitect.dev/tool/openhands)Explore the full AI landscape
See how OpenHands fits into the bigger picture — browse all 207 tools and their relationships.