Cairn as MCP tools: a browser with a memory, for Claude Code, Cursor and Codex.
Wellknown found it in public sources; nobody has proven control of it yet. Claiming takes one click if the repository is under your GitHub account, or a small file on your domain otherwise. Verified owners get the badge, 15-minute checks, status alerts, edits that outrank crawled data, and a ranking boost.
Agents can do it too: POST https://wellknown.network/api/v1/claims with {"agent":"cairn-browser-mcp","method":"well_known_file"} — machine-readable steps at claim.json, guide at /docs/claim.
Everything here was measured by our prober or read from a registry. Nothing is self-reported.
Attributed to the source that supplied each field. Treated as claims, not facts.
# Cairn as MCP tools **Your AI can use websites. But it forgets how, every single time. Cairn makes it remember.** This gives Claude Code, Cursor, Codex or any MCP client a browser that remembers. Your AI walks a site once and Cairn writes down the route; every run after that is **one tool call, no page reading, and zero model calls**. When the site changes, Cairn hands back the single step that moved rather than the whole task. ```bash pip install cairn-browser-mcp playwright install chromium # the browser is a separate download claude mcp add cairn -- cairn-mcp ``` Then just ask, in your own words: *"go to github.com/microsoft/playwright and tell me how many open issues it has."* The first time it explores. After that it is one call. **No API key.** There is no model call anywhere in this package or anything it imports — your AI does the thinking and Cairn supplies the browser and the memory. Optional extra, so one agent can buy a trail from another over x402 on Base: ```bash pip install "cairn-browser-mcp[market]" ``` The engine on its own is [`cairn-browser`](https://pypi.org/project/cairn-browser/). Full documentation, the tool list, and the deletion test that proves the memory is load-bearing: **https://github.com/rohit-jsfreaky/cairn** MIT licensed.
Mapped onto the structured taxonomy from declared text and observed tool names. Confidence shown for derived entries.
Every source is kept verbatim. Field changes are logged as events.