Set it up with your coding agent
You don't have to read all of these docs. Hand them to your AI coding agent — via MCP and Skills — and it will integrate, configure, and build on agent4.io for you.
Don't read the docs. Delegate them.
This documentation is written to be read by machines as well as people. agent4.io ships an MCP server and a set of Skills so your coding agent — Claude Code, Codex, OpenClaw, Hermans, or any MCP-capable assistant — can operate the platform directly.
Point your agent at this doc set and say what you want. It will discover the endpoints, wire up the integration, upload your knowledge, configure agents, and embed the widget — end to end.
The fastest integration is the one you don't write. Give your agent the MCP connection and these docs; review its work instead of authoring it.
How it fits together
What the agent can do for you
- Provision & configure — create agents, set playbooks, connect knowledge bases.
- Ingest knowledge — upload and refresh documents, verify retrieval.
- Wire channels — generate the web embed snippet, set up Telegram/WhatsApp.
- Build on the API — scaffold a custom integration against the REST API (see Chat API) or a white-label app (see White-label & custom apps).
Getting your agent connected
- Create a tenant and API key (see Quickstart).
- Add the agent4.io MCP server to your coding agent's MCP config with your key.
- Hand it the task. The Skills bundled with the platform encode the common workflows so the agent follows the right steps without you spelling them out.
The MCP server and Skills are part of the platform integration surface and are evolving. Exact connection details are documented alongside the MCP endpoint in the console.