Your setup could help someone.
The tools you use. The little habits that help. Let your agent turn them into a page others can learn from. You choose what goes live.
1. Connect your agent
Choose your agent above, then copy this command into your terminal.
codex mcp add mysetup --url 'https://mysetup.ai/mcp'2. Sign in with GitHub
If sign-in doesn’t open automatically, run:
codex mcp login mysetup --scopes setup:read,setup:write,setup:publishSign in with GitHub, then approve the connection. If you’re new here, that creates your account too.
3. Let your agent take the first pass
Paste this into your conversation:
Help me share how I work with AI on mysetup.ai. First read get_setup with target:self and its contributionGuide. Use our existing conversation and relevant local tools within my permissions to discover my agents, harnesses, skills, connections and working practices. Follow the guide’s discovery limits; keep secrets, raw configuration and private project details out of the draft. Explain how my tools work together and develop a concrete workflow where the evidence supports it, rather than listing installed software. Draft from what you know without a questionnaire; ask one short question only if an essential gap prevents a useful draft. If local inspection is unavailable, use our conversation without claiming otherwise. Save a private draft, show me the exact preview, and wait for my explicit approval before publishing.Your agent uses what it already knows and can safely discover to write your first draft. Check the draft, leave out anything private, and share when you’re happy.
Already written it down? Ask your agent to start from your README and link back to GitHub. When your setup changes, just ask your agent to update your page.
Using another MCP client?
Add a Streamable HTTP server at this address, then complete its OAuth sign-in.
https://mysetup.ai/mcpFor contributing, request setup:read, setup:write, and setup:publish.
Need a hand connecting? Codex ↗ · Claude Code ↗ · Pi MCP Adapter ↗