See your site the way AI agents see it.

Check how ChatGPT, Claude, Perplexity and 19 other agents treat any URL.

$ brew install tragent-io/tap/acurl

$ acurl shop.example/products/wool-runnerFetching https://shop.example/products/wool-runner with 22 profiles... PROFILE            STATUS  TIME    RESULT             DETAILSChrome (Browser)   200     412ms   allowed            Wool Runner | Example StoreChatGPT-User       403     51ms    challenged         challenge: CloudflareClaude-User        200     388ms   allowed            OKPerplexity-User    200     402ms   allowed            product: price 89.00 vs browser 98.00OAI-SearchBot      200     97ms    cloaked            31% similar to browserPerplexityBot      200     371ms   allowed            OKGPTBot             200     365ms   robots-disallowed  robots.txt: disallowedGoogle-Extended    -       -       robots-disallowed  robots.txt token (not fetched)Googlebot          200     390ms   allowed            OK AI readiness 71/100 (C)  ✗ AI assistants can fetch the page            16.7/25  ChatGPT-User (challenged)  ✗ AI search crawlers can fetch the page       10.0/20  OAI-SearchBot (cloaked)  ✓ Content is in the HTML without JavaScript   15.0/15  ✓ Product structured data is present          10.0/10 How to improve  1. Allow ChatGPT-User, Claude-User and Perplexity-User in your WAF's verified-bot rules.  2. Serve OAI-SearchBot the same product content as browsers. Run with --diff OAI-SearchBot.

One verdict per agent.

allowed	The agent got the same page as the browser.
blocked	An error status (403, 401, 404, 5xx) instead of the page.
challenged	A WAF or CAPTCHA challenge page.
rate-limited	A 429, with the Retry-After delay.
cloaked	The visible text differs substantially from what the browser got.
robots-disallowed	robots.txt tells this agent to stay out.
error	The request failed: DNS, TLS, timeout or connection reset.

Scripts, CI and AI assistants.

$ acurl -f json shop.example/products/wool-runner
{
  "url": "https://shop.example/products/wool-runner",
  "results": [
    {
      "profile": "ChatGPT-User",
      "category": "ai-assistant",
      "status_code": 403,
      "duration_ms": 51,
      "bot_challenge": "Cloudflare",
      "classification": "challenged"
    },
    {
      "profile": "Perplexity-User",
      "status_code": 200,
      "product": { "source": "json-ld", "price": "89.00", "currency": "USD" },
      "product_issue": "price 89.00 vs browser 98.00",
      "classification": "allowed"
    }
  ],
  "summary": { "total": 22, "allowed": 16, "challenged": 1, "cloaked": 1 },
  "score": {
    "value": 71, "grade": "C",
    "checks": [ { "id": "assistants-allowed", "points": 16.7, "weight": 25 } ]
  }
}

name: AI agent access
on: [pull_request]
jobs:
  acurl:
    runs-on: ubuntu-latest
    permissions:
      security-events: write
    steps:
      - uses: tragent-io/acurl@v1
        with:
          urls: |
            https://staging.shop.example/
            https://staging.shop.example/products/best-seller
          fail-on: blocked,challenged,score<80

# Claude Code
claude mcp add acurl -- acurl mcp

# Claude Desktop, Cursor
{ "mcpServers": { "acurl": { "command": "acurl", "args": ["mcp"] } } }

Commands

acurl <url>	Fetch a URL as every agent, classify each response and score the page.
acurl robots <url>	robots.txt verdict per agent, Content Signals, sitemaps and llms.txt.
acurl batch [file]	Check URL lists, CSV files, stdin or whole sitemaps.
acurl watch <url>...	Re-check on a schedule and alert on changes.
acurl report <file>	Render a saved JSON report as HTML, Markdown, SARIF, JUnit or CSV.
acurl serve	Local web UI and JSON API.
acurl mcp	MCP server for AI assistants.
acurl profiles list \| show	Browse the agent profile catalog.
acurl keygen	Create an Ed25519 key for Web Bot Auth signing.
acurl init	Write a commented `.acurl.yaml` config file.
acurl completion	Shell completions for bash, zsh, fish and PowerShell.

Full command reference