Model Context Protocol (MCP)
Alongside its REST-style HTTP API, mimoLive exposes a Model Context Protocol (MCP) endpoint that lets AI agents control your live production in natural language. MCP is an open standard for connecting language models to external tools; mimoLive implements it over the standard MCP Streamable HTTP transport at a single endpoint, /mcp, served by the same built-in web server (and secured the same way) as the rest of the HTTP API.
Where the HTTP API is designed for scripts and hardware controllers issuing fixed requests, the MCP endpoint is designed for AI clients: after a short initialize handshake, the client discovers the actions mimoLive offers and calls them as structured tools. These cover the same building blocks you automate over HTTP — starting and stopping shows, switching layers and layer variants live, recalling layer sets, adding and removing layers, controlling sources and output destinations, triggering layer/source/filter signals, and driving media playback — plus read-only tools for inspecting the available document, layer, source, and output-destination types.
Point any MCP-capable client (for example an AI assistant that supports MCP) at http://127.0.0.1:8989/mcp to let it query and operate mimoLive on your behalf.