How to make your API docs queryable by AI agents in 60 seconds

Developers spend hours digging through documentation to answer one question. Now their AI assistants — Claude, Cursor, Windsurf — do the same thing, and they hit the same walls: docs that can't be queried, only read.

AgentReady fixes that. You submit your docs URL, we index it, and any AI agent with MCP support can ask it questions and get cited answers grounded in your actual content. No SDK. No embeddings pipeline to build. No prompt engineering.

Why this matters for your docs

When someone installs your SDK and asks Claude "how do I authenticate with the API?", Claude currently either hallucinates an answer or says it doesn't know. With AgentReady, Claude calls ask_site on your domain and gets the real answer — with a link to the exact page.

That's a better experience for your users and a lower support burden for you.

Step 1: Index your docs

Go to agentready.it.com, paste your docs URL, and click Make it agent-ready. We crawl up to 10 pages, chunk and embed the content, and generate a structured summary.

It takes about 60 seconds. Works on any publicly accessible URL — Mintlify, ReadMe, Docusaurus, plain HTML, whatever you use.

Step 2: Connect an MCP client

Add AgentReady to Claude Desktop by editing your config file:

// ~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "agentready": {
      "command": "npx",
      "args": ["-y", "@agentreadyweb/mcp"]
    }
  }
}

Restart Claude Desktop. That's it — no API key, no account needed.

Step 3: Ask questions

Tell Claude to query your site:

"What does docs.yoursite.com say about rate limits?"

Claude calls ask_site automatically and returns a cited answer. Your users get accurate answers. Your docs stay authoritative.

Auto-indexing via MCP

As of last week, ask_site will also auto-index a site it hasn't seen before. So if a user asks Claude about your docs and your domain isn't indexed yet, AgentReady indexes it inline and answers — no portal visit needed.

You can also trigger indexing programmatically:

curl -X POST https://www.agentready.it.com/api/crawl \
  -H "Content-Type: application/json" \
  -d '{"url": "https://docs.yoursite.com"}'

What gets indexed

AgentReady crawls your public pages — up to 10 per site, prioritising your sitemap if you have one. It extracts text content, chunks it for retrieval, and generates a plain-English summary of what the site covers.

JavaScript-heavy SPAs work best if they have a sitemap or server-rendered HTML. Sites behind auth aren't supported yet.

Try it now

Submit your docs at agentready.it.com. Free to try, no account needed. If you run into issues or want to talk about what agent-readiness means for your specific setup, reach out — we're actively building this.