Your team's docs, AI-readable.

The pain

• —Your team docs live in 4 places: Notion, GitHub READMEs, Confluence, half-finished Google Docs. None of them are AI-readable.
• —When a new dev asks the AI "how does our auth work?", it makes something up. Even with paid Copilot Workspace, it doesn't know your codebase's specifics.
• —Building a custom RAG over your docs takes a week. Maintaining it forever.
• —Onboarding doc → "see this page → and this page → and this old wiki → which is wrong now → ask Sarah."

What you do in Memory.Wiki

1. 1
   Import the docs that matter

   GitHub repos of .md files (one URL, Memory.Wiki walks every .md). Notion pages (paste the URL + integration token). Obsidian vaults (drop a .zip). URL pages (any public docs page). Drag-and-drop files from your machine.
2. 2
   Make the hub public

   Settings → toggle Hub public. Memory.Wiki auto-publishes index.md / SCHEMA.md / log.md / llms.txt at the hub root — every AI agent's protocol of choice.
3. 3
   Share the one URL

   memory.wiki/hub/<your-team>. Put it in CLAUDE.md, .cursorrules, .codex-agents — wherever your team's AI loads context from. Each AI fetches and ingests the same way.
4. 4
   Let the lint surface drift

   Needs Review highlights orphan docs (not linked from any other doc) and likely duplicates. Resolve in one click. The KB stays coherent without a docs-czar.

What you get back

• ✓"How does our auth work?" — every AI on your team answers from your real docs, with citations to the source URL.
• ✓New hires read the hub URL on day one and are productive on day two.
• ✓Updates to a doc propagate immediately. The AI fetches fresh markdown every call — no rebuild.
• ✓No custom RAG to maintain. Token cost is visible (token-economy badge on the hub header) and tunable (?compact, ?digest).

Worked example

Try it with what's on your desk right now.

No signup. Drop in your first doc and the URL is yours.

Other shapes the URL takes