Crawlmend documentation

Crawlmend is an AI SEO agent that opens reviewable GitHub pull requests. These pages explain how it works and exactly what it does and does not do.

• Getting startedConnect a domain, GitHub and Search Console, run a crawl, and review your first Crawlmend pull request.
• Supported setupsWhich frameworks, languages and hosting setups Crawlmend works with today, what compile-checked means for each, and what is not supported yet.
• How fixes are generatedFrom finding to pull request: how Crawlmend decides a fix, finds the right file, verifies it and opens a reviewable pull request.
• What data Crawlmend readsThe exact sources Crawlmend reads: your public pages, Search Console performance data, PageSpeed results, repository files, and what is sent to OpenAI.
• Security and permissionsThe access Crawlmend asks for on GitHub and Google, how tokens are stored, how fixes are sandboxed, and how to revoke access.