Cairn — a browser memory for AI agents

Your AI can use websites. But it forgets how, every single time. Cairn makes it remember.

Your AI walks a website once — signing in, clicking, reading — and Cairn writes down the route. Every run after that follows the route instead: one call, no page reading, no model calls at all. When the site changes, Cairn repairs the one step that moved and keeps the rest.

pip install cairn-browser
playwright install chromium        # the browser is a separate download

cairn run --site github.com --task "count open issues"
cairn show github.com              # the route, step by step
cairn forget --site github.com     # make it forget

This is the engine. Most people want it wired into their AI instead, which is cairn-browser-mcp — Cairn as MCP tools for Claude Code, Cursor or Codex.

No API key. Cairn never calls a model; yours does the thinking. Memory is a local SQLite file and needs no account.

Optional extra, for selling and buying trails between machines over x402 on Base:

pip install "cairn-browser[market]"

Full documentation, the design, and the deletion test that proves the memory is load-bearing: https://github.com/rohit-jsfreaky/cairn

MIT licensed.