llms.gist shows you how AI tools describe your product. Audit your URL, see what ChatGPT and Claude say, and get a downloadable fix file that patches the context LLMs are missing.
LLMs answer millions of “what should I use for X?” questions every day. The description they give of your product decides whether you get recommended, ignored, or confused with a competitor. Most of the time, they get it wrong — and no one tells you.
AI coding tools have the same problem. They read your HTML but not your design decisions, rejected alternatives, or boundaries — so they guess. The result is code that looks right but implements the wrong thing.
You give us a URL. We audit your product through every major LLM and score the output across positioning, category accuracy, feature claims, and audience fit. Gaps become a downloadable llms.gist file you can drop into your repo or paste into Cursor, Claude Code, or ChatGPT.
Run an audit any time you ship a meaningful change — a new positioning paragraph, a new feature, a competitor launch — and see how the AI tools your buyers use respond.
The fix file is a .gist, an open format for AI-readable product context. Anyone can author one by hand or with the Claude Code skill. llms.gist produces one as the output of every audit. The spec is MIT-licensed and tool-agnostic.
Pattern references in the fix file link to aiuxdesign.guide, which documents 36 AI UX patterns from shipped products.
Built by Imran. Feedback and spec contributions welcome via GitHub.