llmaudit-mcp

MCP server for llmaudit.app. Lets anyone's assistant measure whether a brand appears when buyers ask an AI for a recommendation.

Full spec: specs/features/geo-13-mcp-server.md in the private repo llm-rank-tracker.

What it exposes

• start_visibility_check starts a measurement and returns a runId.

• get_visibility_check collects it after the fact. Returns running until the providers respond.

• llmaudit://methodology explains how it's measured.

They're two tools and not one because the measurement takes up to two minutes: the audit is 15 to 60s and the measured map about the same.

Related MCP server: ai-visibility-mcp

What it returns, and what it does NOT

It returns the measured map: the buyer's questions consulted live and who got each answer. Never the providers' self-report, which is what each model believes about the brand. The two numbers diverge greatly in practice, and the self-report is always the optimistic one: that's why the tool's contract does not expose it in any field.

It never returns a 0-100 score: a three-band verdict and counts.

Money

Each free measurement consumes llmaudit's API keys, on the order of USD 0.05 to 0.15. Three guards, all verified by tests:

1. One free measurement per domain every 30 days, checked BEFORE calling the API.

2. MCP_MONTHLY_RUN_LIMIT (default 200) as a channel-wide ceiling, separate from the web funnel budget.

3. The rate limit count in prod goes by client identity, not by the IP of this server, which rotates in serverless.

Env

Development

npm run dev     # xmcp dev
npm test        # vitest, sin red
npm run build   # xmcp build

The tests run against real fixtures in fixtures/ (already-paid runs from the prospecting phase). None of them go out to the network or call a provider.