OpenHands Developer Guide
OpenHands is an open-source platform for software-development agents. This guide explains how OpenHands is shaped as a product and codebase, how to get a working local setup, how to think about configuration and safety, and where to contribute.
This guide is based on the official OpenHands docs and the official OpenHands/OpenHands repository, checked on June 26, 2026.
In this section​
- Setup and first run
- Architecture and runtime
- Configuration and security
- Extending and contributing
- Simple user guide
1. What OpenHands is​
The shortest accurate description is:
OpenHands is a development agent platform that can inspect a codebase, run commands, edit files, and iterate toward a software task.
That makes it much closer to Codex CLI, Claude Code, or Goose coding workflows than to document-centric assistants such as AnythingLLM.
2. The mental model​
Think of OpenHands as four cooperating layers:
| Layer | What it does |
|---|---|
| agent runtime | Plans and executes development tasks |
| sandbox or execution environment | Runs commands and edits safely |
| UI and session surfaces | Lets users inspect and steer runs |
| integrations and automation | Connects repos, tasks, and external systems |
3. Why developers would choose it​
- It is focused directly on software development tasks.
- It has both a product experience and a serious open-source codebase.
- It supports local and hosted-style usage patterns.
- The project is explicitly organized around agents that can do more than autocomplete.
4. Where it fits best​
OpenHands is strongest when you want an agent that can actually work through a coding task with tools and feedback loops. It is not the best fit if you only want a simple text chatbot for business knowledge.
5. Recommended reading order​
- this overview,
- introduction and local setup docs,
- architecture and development docs,
- repo structure,
- contribution material before changing runtime behavior.