Developer Documentation

Build on Runnit. This section covers the HTTP API and the Model Context Protocol (MCP) integration for connecting external AI agents. Looking for how to use the product itself, including the built-in AI agents? See the User Guide.

New: the public REST API

Runnit now has a public REST API at /api/v1 covering projects, tasks, organisations, clients, users, reports, schedules, collections, assets, and search. Create a personal API key at Admin → API Keys and start with the REST API Overview.

Public REST API

REST API Overview The /api/v1 surface: discovery endpoint, resources, and example requests.

API Keys & Authentication Create per-user API keys and send them with REST and MCP requests.

API Reference

Image Generation API

Video Generation API

File API

Asset Analysis API

Agent Schedule API

Projects API

MCP

MCP Server Connect agents over the Model Context Protocol with user-scoped tokens.

MCP Tools The tool contract, authentication modes, and the read and write tool groups.

Reading Data Retrieve projects, users, tasks, clients, and more via the MCP read tools.

Looking for agent building?

Runnit’s AI agents run inside the product, so there is no external agent-hosting integration to build. Administrators create and customise agents from the app; see the Agents Catalog. External tools connect through the MCP server.