Get results where your team already works

Run the site check and fetch paste-ready fixes without leaving your editor or chat. It's free and needs no plan. It doesn't return a Score (that's the full check at ruleriq.com/grader). It works through our MCP server, a Streamable HTTP endpoint (JSON-RPC 2.0) at https://ruleriq.com/api/mcp.

• ruleriq_audit_domainRun RulerIQ's free site check on a domain: Site readability 0–100 (crawler access, text in HTML, structured data, llms.txt, basics), what we found on the site, and paste-ready fixes. It does not query ChatGPT, Perplexity, Gemini or Claude and returns no Score. The full check runs at https://ruleriq.com/grader.
• ruleriq_gsc_blindspotsTakes Search Console queries you paste in (query, impressions, clicks, position), turns them into buyer questions and lists the ones where you rank on Google page 1. It makes no network calls and does not query AI engines: AI mention and citation rates are returned as 0, not measured.
• ruleriq_generate_fix_kitReads a site's public pages and returns paste-ready fixes: FAQ HTML block, fact comparison table, JSON-LD (SoftwareApplication + FAQPage) and llms.txt text. Nothing is changed on the site, and no Score is returned.

With a personal access key, these read tools show your own tracked brands: Score, history, fix list and answers. Paid plans are not open yet; once you track a brand you create a key under Settings, Keys and connections. Without a key they are not listed. Keys read by default; tick Write access when you create one and ruleriq_mark_fix_done is added, which marks your fixes done or undone.

• ruleriq_list_brandsLists the brands this account tracks: domain, id, plan, status, last check and next scheduled check. Use it first to get a brand's domain or id.
• ruleriq_get_scoreThe latest RulerIQ Score (0-100) for a tracked brand with its margin of error, the four components (Named, Cited, Facts right, Site readability), the week-over-week verdict and the trend. The sentences returned are computed by RulerIQ: quote them rather than restating changes in your own words, and treat a change as real only when the sentence says so.
• ruleriq_get_historyWeekly check history for a tracked brand: the same rows as the History page and its CSV (Score, margin of error, change vs last week, verdict), oldest first, capped at 104 rows. Optional filters: from/to dates, one engine, one question (adds runs naming you and best rank).
• ruleriq_get_fix_listThe fix list from the latest check, ranked as in the app (wrong facts, then questions where the brand is not named, then pages to get onto). Each item has a status (open, improving, verified, regressed, not_enough_data) and a sentence from RulerIQ's verified-fix check. This tool only reads; mark an item done with ruleriq_mark_fix_done (needs a key with write access).
• ruleriq_get_answersPer question and engine in the latest (or a given) check: whether the brand was named, agreement across runs (e.g. 3/5), best rank, plus the pages the engines cited in that check. RulerIQ does not store the engines' answer text. Question wording is returned for checks made since 2026-11-04 (earlier checks are numbered).
• ruleriq_mark_fix_doneMarks one fix-list item done (done=true, the same as the Mark done button in the app) or undone (done=false, back to still missing). Needs a personal access key created with write access; read-only keys cannot call it. The fixId comes from ruleriq_get_fix_list, and only tracked brands owned by the key's account can be written. A fix only counts as verified after at least two later full checks confirm it, so the item keeps showing 'not enough data' until then. (Write-access key only.)

claude mcp add --transport http ruleriq https://ruleriq.com/api/mcp

{
  "mcpServers": {
    "ruleriq": {
      "url": "https://ruleriq.com/api/mcp"
    }
  }
}

{
  "servers": {
    "ruleriq": {
      "type": "http",
      "url": "https://ruleriq.com/api/mcp"
    }
  }
}

codex mcp add ruleriq --url https://ruleriq.com/api/mcp

[mcp_servers.ruleriq]
url = "https://ruleriq.com/api/mcp"

https://ruleriq.com/api/mcp

Add the RulerIQ MCP server to my MCP config. It's a remote Streamable HTTP server at https://ruleriq.com/api/mcp with no authentication. Then use its ruleriq_audit_domain tool to check example.com.

Use RulerIQ to check example.com

Full setup notes, including how to test from a terminal, are in the MCP docs.

Slack and Notion connect in one click from your account (with Linear, which uses a key you paste). They are in beta and only for tracked brands, and paid plans are not open yet, so join the waitlist to be invited. Email reports are planned. Today the free grader gives you the report and paste-ready fixes in your browser.

Start from the searches your customers really make. Two ways: paste a CSV export ( POST /api/gsc), or connect read-only via OAuth (planned for the paid plans). We never hold write access.

• /openapi.jsonOpenAPI 3.0 spec of every endpoint.
• /agent-facts.jsonMachine-readable facts, pricing, engines, MCP tools, and endpoints.
• /llms.txtRulerIQ's own llms.txt: a plain-text summary of this site for AI assistants.

See also the methodology and how we handle data.