Use this site with AI

Everything on this site is public, and I want it to be easy for the AI tools you already use to read it, search it, and cite it properly. There are four ways in, from most to least interactive.

MCP server

The site runs a read-only Model Context Protocol server. Point an MCP-capable assistant at it and it can browse and search everything here, rather than guessing from search results.

Server URL: https://chrischinchilla.com/mcp (Streamable HTTP, no sign-in)

To add it:

Claude Code: claude mcp add --transport http chrischinchilla https://chrischinchilla.com/mcp
Claude (web and desktop): add a custom connector in your connector settings and paste the URL above.
Other clients: most accept a remote HTTP server URL in their MCP settings. The MCP Inspector is handy for trying it out.

Tools it offers:

Tool	What it does
`list_content_types`	Lists every content type (blog, books, podcasts, videos, music, games, and more) with counts, plus the four site categories: tech, writing, music, and games.
`list_content`	Browses a listing, newest first, for one content type or all of them. Filter by category, tag, or date, and page through the results.
`search_content`	Fuzzy search across titles, summaries, and tags, with the same filters.
`get_content`	Fetches one entry in full, with its Markdown text and links, by type and id or by page path.
`list_tags`	Lists the tags in use and how many entries use each.

Try asking: “Using the chrischinchilla.com MCP server, what has Chris written about podcasting?” or “List Chris’s most recent music releases.”

In browsers that support WebMCP, an early web standard that lets a page offer tools to the browser’s built-in AI agent, every page here registers the same tools directly. There is nothing to install or configure. It also adds a tool for reading any page on the site, such as the CV. Browser support is still limited, so if your browser doesn’t have it, the page simply behaves as normal.

JSON content API

The same data the tools use is available as plain static JSON:

/api/content/index.json lists every entry with its type, summary, date, tags, and categories.
/api/content/<type>.json lists one content type, for example /api/content/books.json.
/api/content/<type>/<id>.json returns a full entry, including the Markdown text. Articles that have a video version also include a videos list with the YouTube link, chapters, and a transcript where one exists (auto-generated transcripts are marked as such). For those videos, the entry in the videos listing points at the article, which is the canonical page.

llms.txt

If your tool prefers a single text file, /llms.txt is an index of everything on the site, and /llms-full.txt contains the full text. Each article also has an “Open with AI” menu in the bottom right that can copy it as Markdown or open it in an assistant.