Make a page ready for search engines to index and AI answer engines to quote.
Generate the machinery a page needs to be read by both: robots.txt that welcomes AI crawlers by default, sitemap.xml, a page <head> with JSON-LD, and llms.txt.
Audit a live page and get two scores instead of one. Findable asks whether an engine can reach and index the URL. Citable asks whether, once reached, the page can be extracted and attributed. Two scores mean the result tells you which problem you have, not only that you have one.
The audit reports what it observes on one pass: the page's robots meta tags, the X-Robots-Tag header, an expired unavailable_after, the origin's robots.txt, and whether the canonical target answers. It cannot know which crawlers honor a directive aimed at one of them, so it reports the block it saw and never reports who is still indexing you. audit_html scores markup before deploy, so it checks a template, not a deployment.
All six operations run over MCP: audit_page, audit_html, generate_head, generate_robots, generate_sitemap, generate_llms_txt.