OpenHands (formerly OpenDevin) is an open-source autonomous AI software development agent and orchestration platform. It provides an extensible runtime environment that enables AI agents to inspect repositories, write and modify code, execute terminal commands, run tests, and interact with external developer tools. Developers interact with OpenHands through multiple interfaces, including a terminal CLI, the Agent Canvas desktop application, an SDK, and managed cloud environments. OpenHands can run its native autonomous agent architecture or orchestrate external CLI agents such as Claude Code, Codex, and Gemini CLI. It connects to a wide range of LLM providers (including Anthropic, OpenAI, and Google) via standard API keys or local inference endpoints. Unlike standalone terminal assistants, OpenHands operates as a full agent execution platform and control plane capable of running background automations. It integrates with over 70 third-party platforms—including GitHub, Linear, Jira, Slack, and Snyk—and supports the Model Context Protocol (MCP) to dynamically connect agent skills and context sources. OpenHands is open-source software with an active community on GitHub, complemented by OpenHands Cloud and Enterprise tiers that offer centralized control planes, team governance, and managed infrastructure.
Agent Canvas Interface — Desktop and terminal-accessible interface for tracking multi-turn agent conversations, tool outputs, and execution states. - Bring-Your-Own-Agent Architecture — Execute native OpenHands autonomous agents or orchestrate external CLI agents such as Claude Code, Codex, and Gemini CLI. - Prebuilt Workflow Automations — Trigger autonomous agent sessions on events across GitHub pull requests, Linear issues, Jira incidents, and Slack mentions. - Model Context Protocol (MCP) Support — Connect external tools, context servers, and custom agent skills via standardized MCP interfaces. - Extensive Tool Integrations — Connect directly to over 70 developer and monitoring tools including Snyk, Supabase, Sentry, Datadog, and Notion. - Flexible Deployment Models — Run locally via npm and Docker, integrate programmatically via the Python SDK, or deploy via OpenHands Cloud and Enterprise control planes.
Use cases
Use Case: Automated CI Failure Remediation Configure OpenHands to listen for failed GitHub Actions workflows, download and inspect the execution logs, reproduce the failure locally in its sandbox, apply the fix, and open a remediation pull request.
Use Case: Security Vulnerability Patching Connect OpenHands with tools like Snyk to automatically ingest vulnerability reports, analyze the affected dependencies or source files, update the codebase, and verify that test suites pass before submitting patches.
Use Case: Slack and Issue Tracker Automation Monitor Slack channels, Linear tickets, or Jira incidents to triage incoming bug reports, gather recent commit logs, reproduce issues, and post drafted solutions or retrospectives back to the team.
Who it is for
OpenHands is built for software engineers, DevOps teams, and platform engineering organizations that need an open-source, customizable agent platform to automate development workflows across their existing toolstack. It is not intended for developers seeking a lightweight, inline IDE autocompletion plugin with no agentic execution or sandboxing requirements.
Install guide
Installation Install the OpenHands Agent Canvas CLI globally using npm: bash npm install -g @openhands/agent-canvas Alternatively, clone the open-source repository for local development and containerized execution: bash git clone https://github.com/OpenHands/OpenHands.git cd OpenHands
First Run Launch the Agent Canvas interface from your terminal: bash agent-canvas Configure your LLM provider API keys during initial startup or export them in your shell environment: bash export ANTHROPIC_API_KEY="your-anthropic-api-key" # Or for OpenAI: export OPENAI_API_KEY="your-openai-api-key"
Example Session Start an OpenHands task to fix failing tests in a repository: bash agent-canvas --prompt "Analyze the failing tests in tests/test_auth.py, resolve the token expiration issue, and verify tests pass."
Plandex — Plandex is an open-source terminal-based AI coding agent designed to handle complex, multi-file engineering tasks in large codebases. Created by…
OpenCode — OpenCode is an open-source AI coding agent developed by Anomaly that operates across terminal environments, desktop applications, and IDE extensions.…
Subtask — Subtask is an open-source CLI agent harness and orchestration tool developed by zippoxer in Go. It is built to coordinate…
pi-reflect — pi-reflect is an open-source agent harness and self-improvement extension for pi coding agents, developed by jo-inc. It inspects agent session…
EchoCoding — EchoCoding is an open-source agent harness and audio-augmented developer tooling layer created by launsion-boop. Designed to run alongside AI coding…
OSOP Agent Rules — OSOP Agent Rules is an open-source orchestration harness and rule repository developed by Archie0125 (osop.ai) designed to standardize agent behavior,…