OpenHands/OpenHands

Project overview

OpenHands is a self-hosted developer control center for starting, managing, and automating coding agents across local and remote environments. It supports running agents across multiple backends and allows developers to bring their own LLM.

Project type: AI Agent · AI Coding
Use cases: Coding & Development · Automation
Deployment: Refer to project documentation
License: License pending

Best for

Developers who need to manage and automate coding agents across local and remote environments using their own LLMs.

Key capabilities

Run coding agents locally, in Docker, on VMs, or in the cloud, switching between backends without losing focus.
Use any LLM with the coding agents.

Limitations and risks

Running the agent server directly without a Docker sandbox gives the agent full access to your filesystem.
Running without a sandbox exposes your filesystem to full access by the agent.

Getting started

Installation is considered easy and can be accomplished via a single npm command or Docker container without writing code. Users run npm install -g @openhands/agent-canvas, start agent-canvas, and access the UI at http://localhost:8000.

Evidence and sources

README: status-beta-blue
README: It's a developer control center for starting conversations and automating everyday tasks
README: Switch between local, remote, and cloud agents without losing focus
README: Create automations and workflows that integrate with Slack, GitHub, Linear, and more. Run on a schedule or in response to webhook events
README: Use with OpenHands, Claude Code, Codex, Gemini, or any agent with Agent-Client Protocol (ACP).

Reviewed at: 2026-07-17T04:55:55Z