CrawlSpecBot/1.0

SEO audits your agents can act on.

Crawl to spec in one pass. A card closes itself when the next crawl confirms the fix - implementation-ready recommendations, not another dashboard to read.

34 checks cross-mapped to WebSite Auditor, parser-exact on a real 1,376-finding export. Read the calibration.

Start on Founding 50See the loopEmail us for a demo

Demo requests go to support@crawlspec.dev

Or see it on your own site, free

25 pages, no account, no card. What the free scan does

25 pages, depth 2, robots.txt honoured. Most scans finish in under a minute, and the report link stays live for 14 days.

GET /v1/specs/4821
{
  "issue": "missing_canonical",
  "severity": "critical",
  "affected": 342,
  "template": "product-detail",
  "action": "add_canonical_tag"
}
Issue catalog

34

Persistent checks, severity presets

MCP tools

11

Native agent access, no scraping

Audit output

Spec

Template-level fixes, not raw errors

The loop closes itself

Your agent reads the spec through MCP and ships the fix in your repo. The next crawl re-observes the template. The card closes itself.

CrawlEvery URL, politelyClusterGrouped by templatePrioritizeRanked, ready to shipAgent ships fixIn your repo, not oursRe-crawlChecked, not assumedClosedCloses on re-observe

The content half runs on the same evidence. Search Console queries and the keyword map feed topic clusters; a brief is generated from what the crawl already knows about the template, not from a blank page. Your writer or your agent ships the page and marks the brief applied - the one write the Spec API allows.

Built for the agent era

A machine-readable Spec API and a 11-tool MCP server expose audits, recommendations, briefs, topics, keywords, page audits, and cannibalization. Content pipelines start from a spec instead of re-researching keywords.

MCP over the Spec API - self-hosted, key-gated

tools/list

  • list_projects
  • get_recommendations
  • get_briefs
  • get_topics
  • get_keywords
  • get_consolidated_audit
  • mark_brief_applied
  • get_page_audit
  • get_cannibalization
  • get_brief_prompt_pack
  • request_brief_recheck

Not another enterprise suite

Crawl-to-spec, API-first, agent-native. Pro is $39/mo, or $195/yr on Founding 50, and the Spec API and the MCP server are in every paid plan rather than an add-on. Three things decide whether this is the right instrument: the output is machine-readable because that is what it was built to emit, the loop closes on re-crawl instead of on somebody ticking a box, and the calibration behind the numbers is published in full.

The app never edits or publishes anything on your site. Every brief is an editable artifact you arrange and finish, then hand to your team or your own AI tools.

Spec API and MCP server ship day one, not roadmap.

No backlink index. Crawl-to-spec only - not a substitute for a link-graph tool.

No rank tracking. Search Console data flows in; a daily rank-position tracker does not.

Founding 50

The first 50 workspaces to subscribe on the yearly Pro price pay $195/yr instead of $390/yr - 50% off, locked in for as long as the subscription stays active.

Start on Founding 50What Founding 50 buys, and what renewal does

33 / 33 categories parser-exact on a 1,376-finding export