Getting started
Crawlmend is an AI SEO agent that crawls your site, reads Search Console and PageSpeed evidence, and opens a reviewable GitHub pull request for each fix. This page walks through setup from sign-up to your first pull request.
1. Create an account
Sign up with your email and a password. Crawlmend sends a confirmation link, and you cannot sign in until the email is confirmed.
2. Add your site
Enter the domain you want to improve. Crawlmend only crawls the host you enter, and it refuses private or internal network addresses.
3. Connect GitHub and pick the repository
Install the Crawlmend GitHub App on the repository that builds the site and authorise it, then choose that repository. Crawlmend detects the framework from the repository (see supported setups) so fixes land in the files your framework actually uses. Details of what the app can access are in security and permissions.
4. Connect Search Console (recommended)
Connecting Search Console gives Crawlmend read-only performance data: queries, pages, impressions, clicks, click-through rate and position. That data is what turns a generic finding into an evidenced one, for example a page with many impressions and a weak title. Crawling and technical fixes work without it.
5. Run a crawl
The crawl checks every reachable page for technical issues (titles, descriptions, headings, canonicals, structured data, hreflang, JavaScript-dependent content), plus site-wide checks such as robots.txt, sitemap.xml, llms.txt and whether AI answer bots can fetch your pages. Findings appear as opportunities with the evidence next to them.
6. Review a fix and open the pull request
Open an opportunity to see the proposed change as a diff: the file, the before and the after. Approve it and Crawlmend opens a branch and a pull request on GitHub. Nothing merges automatically. Your team reviews it like any other change, and closing the pull request discards it.
See how fixes are generated for what happens between the click and the pull request.