Agents
Everything an agent needs to work in a Treehouse workspace, and everything you need to connect one.
Treehouse is built so an agent can work in a workspace as naturally as a person can. There's no integration to write: agents use the files, through a synced folder or the Treehouse MCP server, and every change they make is versioned and attributed.
If you're an agent, start with the agent handbook. If you're connecting one, start with how agents work with Treehouse, then follow the page for your client. Once it's connected, the playbooks show how to shape a workspace so agents are genuinely useful in it.
| What | Where |
|---|---|
| MCP server | https://api.trytree.house/api/mcp (Streamable HTTP) |
| REST API | https://api.trytree.house/api |
| Authentication | An API key in the x-api-key header. Authorization: Bearer is not accepted. |
| Agent skill | https://trytree.house/.well-known/agent-skills/treehouse/SKILL.md |
| These docs as markdown | Add .md to any page URL, or read /llms.txt |
In this section
- Connect an agentGive Claude Code, Codex, Cursor or any other agent access to a workspace, with its own key when you want its work labelled.
- Agent handbookWritten for agents. How to reach a Treehouse workspace, the rules that keep everyone's work safe, and the conventions the web app understands.
- API referenceThe protocol reference - authentication, every MCP tool, the REST API and the change feed.
Last updated