The Agentic Readiness Check, right inside Claude.
ARC is an MCP server. Add it to Claude and just ask: “Can AI agents buy from this store?” ARC runs AI shopping agents modeled on the real surfaces (ChatGPT, Claude, Google AI Mode…) to try, then hands back the score, the blocker, and the fix — without you ever leaving the chat.
We're agent-native. The only question is whether your store is.
▸ you: can AI agents actually buy from nike.com?
claude → calling ARC · scan_store(nike.com)
claude → ARC · check_scan…
All without leaving the chat.
Three clicks and you're live.
If you use Claude, you can add ARC in under a minute — right from the app. No command line, nothing to install.
Open Claude settings
In Claude, click your name (bottom-left) → Settings.
Add a connector
Open Connectors → Add custom connector.
Paste the URL
Name it ARC, paste the URL below, click Add.
https://www.arcreport.ai/api/mcpThat's it. Now just type in Claude: “Use ARC to check if agents can buy from nike.com.”
▸ PREFER THE TERMINAL? DEVELOPER SETUP (CLAUDE CODE, CURSOR)
Claude Code
claude mcp add --transport http arc https://www.arcreport.ai/api/mcpCursor / any MCP client
{
"mcpServers": {
"arc": { "url": "https://www.arcreport.ai/api/mcp" }
}
}Two tools. Plain English.
Runs AI shopping agents (modeled on the real surfaces) at a store and starts a readiness scan.
Returns the ARC Score, how many agents could buy, the blocker, and the fix.
Put agent-readiness where your agents already are.
Connect ARC to Claude, or run a scan on the web first to see what you get.