Agent Ready

Scan any public URL for AI agent readability and get back a 0–100 score plus per-check remediation hints. The same engine that powers [agent-ready.dev](https://agent-ready.dev), packaged as an MCP server for Claude Desktop, Claude Code, Cursor, VS Code, and Windsurf. 60 checks across 4 spec families: - 38 against the Vercel Agent Readability Spec (15 site-wide + 23 per-page) - 10 against the llmstxt.org specification - 12 against agent-protocol manifests: MCP Server Cards (SEP-1649 / RFC 9728), A2A Agent Cards, Wildcard agents.json, agent-permissions.json, UCP, x402 Payment Required, NLWeb **Tools:** `scan_site`, `get_scan`, `ask` **Prompts:** `scan`, `interpret_scan`, `remediation_plan` **Skill:** `skills/agent-ready/SKILL.md` for Claude routing Scanning (`scan_site` / `get_scan`) requires an Agent Ready Pro API key — issue one from [the dashboard](https://agent-ready.dev/dashboard/api-keys). `ask` (natural-language search over the methodology, checks, and specs) is public and needs no key.

First seen 2 Oct 2026. Evidence as of 2 Oct 2026.