# Nimo for AI coding assistants

Bring Nimo audits, performance history, and follow-up comparisons into your AI coding assistant through MCP.

Canonical URL: https://heynimo.com/for/coding-assistants
Markdown URL: https://heynimo.com/for/coding-assistants.md

## Give your agent eyes for speed

Nimo is an AI website performance agent. Use saved audits, lab diagnostics, available field data, and performance history to decide what to inspect first.

## Workflow

1. Start with evidence: ask your assistant to read the latest audit for the page and device you are investigating.
2. Choose the next fix: ask for a likely cause and a focused change. Review code changes in your coding assistant.
3. Check what changed: after deployment, run a matching audit and compare it with the baseline. Keep immediate lab results separate from slower field trends.

## Clients and setup

Claude Code, Cursor, and ChatGPT can connect through MCP using their supported client setup. Availability and configuration depend on the client and account. Pi needs an MCP extension; MCP is not built into Pi's core.

[Set up Nimo MCP](https://heynimo.com/docs/mcp)
[Pi extension documentation](https://pi.dev/docs/latest)

## Questions to ask

- Which monitored page should I investigate first?
- Why do the lab result and field data disagree?
- Compare my last two mobile audits. What changed?

## Limits

Account evidence requires your Nimo connection. New audits use your account allowance. Missing field data stays unknown, not a passing score. A follow-up audit checks the deployed page; it does not guarantee a field improvement or search ranking. The page's editor graphic illustrates the workflow, not a live result.