CervellaSwarm MCP server

CervellaSwarm is a helper for Claude Code that gives your AI a small team of specialists to pass work to. There is one planner, three checkers, and twelve workers for things like backend code, frontend, testing, docs, and security. It is handy if you want a second pair of eyes on the code your AI writes, without setting up anything complicated.

What this MCP server does

You ask your AI for something, like writing a login endpoint or reviewing a file. Your AI uses this helper to pick the right worker for the job, for example the backend worker for code or the tester for checks. The helper sends the task to that worker using your own Anthropic API key, and the worker does the work. You get the result back in the chat, and for some tasks a quality guardian reviews it before you see it. You can also ask which workers exist, check that your setup is correct, and see how much of your quota you have used.

Click to zoom

What you can do with it

• Hand a coding task to a specialist worker like backend, frontend, or tester
• Ask a guardian to review code quality, research accuracy, or deploy safety
• Ask the architect to plan a complex task before any code is written
• List all 17 workers and what each one is good at
• Check that your API key and configuration are set up correctly
• Check how much of your usage quota is left

Try asking your AI

• “Use spawn_worker to create a REST API endpoint for user authentication”
• “Ask the tester worker to write tests for my login function”
• “Have the security worker review this file for vulnerabilities”
• “Show me the list of available workers and their specialties”

What it gives back to you

You get the worker's answer back in your chat, usually as text: code, a plan, a review, or a short report. The list_workers tool returns a simple list of names and specialties. check_status and check_usage return short answers about your setup and quota. Nothing is written to your files unless you ask your AI to do that afterwards.

Before you start

Who it's for

Developers and technical teams who use Claude Code and want a structured way to split work between specialists and have it checked.