Overview

AI agents like Claude Code, Cursor, and Codex can do more than answer questions — connected to the right tools, they take action on your behalf. The Model Context Protocol (MCP) is the open standard for those connections: an MCP server offers a set of capabilities, and any MCP-compatible agent can connect to it and use them.

The Eventum MCP server is such a server: it gives your agent everything it needs to work with Eventum. Instead of learning Eventum's configuration and writing generators by hand, you describe the data you want and let the agent build it.

For example, you might ask for:

a week of Apache access logs with a traffic spike at noon, mostly 200s but a burst of 500s during the spike

The result is a working generator that produces exactly that. See how it works for the loop that builds and checks it.

No model required

Eventum has no language model of its own, no API keys, no per-token cost — your agent already brings the model. The MCP server adds the tools to discover, build, check, run, and operate generators:

• Discover — what Eventum can do and how to configure it, with worked examples to start from.
• Build — create, edit, and remove generators.
• Check — confirm a generator works, by validating it and previewing its output.
• Run — run a generator to its configured outputs.
• Operate — start, stop, and manage generators on a live server.

Two ways to connect

Both give your agent the same tools to build, check, and run generators. HTTP adds management of the generators running on a server.

Read next