For AI Agents & LLMs

How to consume the Wildcat docs as an AI agent — Markdown endpoints, llms.txt, the ?ask API, and the Wildcat skill.

This documentation is built to be machine-readable. If you're an AI agent — or building one — here's how to consume it efficiently.

Markdown and indexes

• /llms.txt — a curated, sectioned index of the entire documentation, with a short description of every page. Start here to navigate.

• /llms-full.txt — the full text of the documentation concatenated into a single file, for wholesale ingestion.

• Markdown for any page — append .md to any page URL to get clean Markdown without site chrome (for example, terminology.md).

Ask the docs a question

Appending ?ask= to a page URL opens the documentation assistant with your question pre-filled:

https://docs.wildcat.finance/using-wildcat/terminology?ask=How does a withdrawal cycle work?

This is a browser feature, not an API. The parameter is handled client-side, so a plain GET returns the same HTML as the page without it, and appending ?ask= to a .md endpoint returns 404. If you are an agent fetching over HTTP rather than driving a browser, use the Markdown endpoints and indexes above instead.

The Wildcat skill

For agents that need a single, self-contained briefing on the protocol — concepts and terminology, borrower and lender workflows, contract internals and addresses, and security/legal pointers — load the Wildcat skill (SKILL.md) maintained in the documentation repository under skills/wildcat-protocol/.

Where to start

• New to Wildcat? Read The Elevator Pitch, then Terminology.

• Using the protocol? See Borrowers and Lenders.

• Auditing or integrating? See Security/Developer Dives and Contract Deployments.