Skip to content
RookAtlas
On every plan, Free included

Your AI visibility, inside the tools you use

Connect Claude or another MCP client to the brands you choose, or call the REST API from your own scripts. Read freely; anything that spends credits quotes first.

  • All eight AI engines on every paid prompt run
  • A 95% interval on visibility and citation rates
  • Every run priced before it starts

Opening soon: sign-up opens with live measurement, after our provider pilot.

Your assistant, connected
How visible were we in AI answers this month, and how sure is that?
rookatlas · get_metrics

JBL was named in 47% of 238 answers across eight engines this month. The 95% interval is 30–66%: wide, because 8 prompts is a small sample.

Perplexity names you most (56%), Grok least (38%). Want a quote for a run with more prompts to narrow it?

RookAtlas has a REST API described by an OpenAPI document and an MCP server for AI assistants. Both read brands, prompts, the answers themselves, metrics with intervals, competitors, cited sources and runs, and both start runs only through a quote the caller confirms. API keys carry read, write and run scopes and can be locked to one brand; an MCP connection is granted to one workspace and the brands you pick.

Measured on eight engines

  • ChatGPT
  • Gemini
  • Perplexity
  • AI Overviews
  • AI Mode
  • Claude
  • Grok
  • Muse
01

Ask, don't export

Ask your assistant how visible you were this month, which sites the engines cite, or to quote a run. It answers from the same estimator as the dashboard, intervals included.

  • One workspace and the brands you pick
  • Write and run chosen by you, like an API key
  • Quotes before anything spends
Connect in one line
  • Ask“Which sites do the engines cite most for our prompts?”
  • Ask“Which prompts does Apple win?”
  • Ask“Quote a run of our weekly prompts, then start it.”
Claude Code command
claude mcp add --transport http rookatlas https://app.rookatlas.com/api/mcp
02

A REST API with an OpenAPI document

Brands, prompts, answers, metrics, competitors, cited sources, runs and credits, described in OpenAPI 3.1. Keys act with their creator's current role and stop working when that person leaves.

  • Read, write and run scopes
  • Keys locked to one brand if you like
  • Runs started only with a confirmed quote
Metrics through the REST API
Request
curl https://app.rookatlas.com/api/v1/brands/{brand_id}/metrics \
  -H "Authorization: Bearer $ROOKATLAS_API_KEY"
Response, abridged
{
  "current": {
    "answers": 238,
    "prompts": 8,
    "mention_rate": {
      "value": 0.475,
      "low": 0.295,
      "high": 0.661,
      "answers": 238,
      "confidence": "low"
    }
  },
  "engines": [ … ]
}
03

For developers: a skill and a CI check

A Claude Code skill teaches your coding agent the API. A GitHub Actions step calls a free endpoint that judges robots.txt for every AI crawler before it ships.

  • SKILL.md served by the app
  • POST /api/v1/audit/robots-txt, free
  • Also an MCP tool, audit_robots_txt
A robots.txt check in CI
.github/workflows/ai-crawlers.yml
name: AI crawler check
on:
  pull_request:
    paths: ["public/robots.txt"]
jobs:
  robots:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Check robots.txt against AI crawlers
        run: # POST https://app.rookatlas.com/api/v1/audit/robots-txt
AI crawler checkFailedpublic/robots.txt: PerplexityBot is blocked by Disallow: /. Perplexity can't cite these pages.

How it works

  1. Step 1

    Connect

    Add the MCP server to your client, or create an API key.

  2. Step 2

    Read

    Metrics with intervals, competitors, cited sources and the answers themselves.

  3. Step 3

    Run, with a quote

    Ask for a quote, confirm it, and the run starts.

What's inside

  • MCP server

    Current MCP specification, OAuth with client metadata documents.

  • REST API

    Everything the dashboard reads, in OpenAPI 3.1.

  • Scoped keys

    Read, write and run, optionally locked to one brand.

  • Quotes before spending

    No call spends credits without a confirmed quote.

  • Claude Code skill

    Your coding agent learns the API from one file.

  • In-app assistantcoming

    Ask about a brand from inside the app, from measured data only.

What it can't do

The API and MCP server read what the dashboard reads, through the same permission checks, and nothing more. Requests are rate-limited per key, and a key acts with its creator's current role.

All features

Questions

Prices, credits and plans are on the pricing page .

Which MCP clients work?

Clients that follow the current MCP specification and identify with a Client ID Metadata Document, such as Claude Code. The connect command is one line.

Does reading through the API cost credits?

No. Reads are free on every plan, Free included. Starting a run costs what it costs in the app, through a quote you confirm.

Is the API on the Free plan?

Yes. The REST API, the MCP server and webhooks are on every plan, rate-limited and credit-gated like the app.

Pick a plan. Every engine is included.

MCP server and API is on every plan, Free included. One credit runs a prompt on all eight engines (Free's five monthly credits run on ChatGPT, Gemini and AI Mode), and every run shows its price before it starts. Compare the plans and work out your credits.

Free tool, usable today: llms.txt generator and validator

Opening soon: sign-up opens with live measurement, after our provider pilot.

Plans from
$36/mo
Engines on every paid prompt run
8
Credits per prompt
1