MCP Server Documentation

Quickstart

1. Open Claude Desktop Settings

Go to Settings → Connectors (or Settings → MCP Servers).

2. Add Custom Connector

Click "Add custom connector" and enter:

3. Authorize

Click "Add" → a browser window will open → log in with your IATO account → click "Authorize". That's it! All 77 MCP tools will appear in Claude Desktop.

{
  "mcpServers": {
    "iato": {
      "type": "streamable-http",
      "url": "https://iato.ai/mcp",
      "headers": {
        "Authorization": "Bearer iato_your_api_key"
      }
    }
  }
}

# Install
python3 -m venv mcp_venv
source mcp_venv/bin/activate
pip install -r requirements-mcp.txt

{
  "mcpServers": {
    "iato": {
      "command": "/path/to/crawler/mcp_venv/bin/python",
      "args": ["-m", "mcp_server.server"],
      "cwd": "/path/to/crawler",
      "env": {
        "IATO_BASE_URL": "https://iato.ai",
        "IATO_API_KEY": "your-api-key"
      }
    }
  }
}

Architecture

┌─────────────┐  Streamable HTTP   ┌──────────────────┐              ┌─────────────┐
│  AI Agent   │ ◄────────────────► │  MCP Server (83) │ ◄──────────► │  IATO API   │
│  (Claude)   │  iato.ai/mcp       │  Python / httpx  │   REST/JSON  │  FastAPI    │
└─────────────┘  (or stdio local)  └──────────────────┘              └──────┬──────┘
                                                                           │
                                                                    ┌──────▼──────┐
                                                                    │    MySQL    │
                                                                    └─────────────┘

Response Format

Every tool returns a consistent JSON envelope:

{
  "success": true,
  "data": { ... },
  "summary": "Created workspace 'My Site' (ID: 42)",
  "next_actions": [
    { "tool": "start_crawl", "args": {...}, "description": "Begin crawling" }
  ]
}

Error responses include recovery_actions with specific tool calls — agents never hit dead ends.

Tool Reference

All 105 tools organized by module. Method badges: GET POST PUT DEL CLIENT

Workspaces (5 tools)

Crawls (5 tools)

Sitemaps (3 tools)

Sitemap Nodes (7 tools)

Content (3) · Changes (2) · Export (2)

Taxonomy (10 tools)

Navigation (4) · Schedules (7) · Recrawl (2)

Analytics (5) · SEO Audit (5) · AI Suggestions (4)

Webhooks (7 tools)

Events: crawl.started crawl.completed crawl.failed changes.detected and more.

Agent Protocol (5 tools)

Extraction Rules (6 tools)

Create and manage CSS, XPath, or regex extraction rules to pull structured data from crawled pages. Rules can be tested against live URLs before saving, then attached to crawls or schedules via extraction_rule_ids.

Workflow Examples

Crawl & Map a Website

create_workspace(name="Client Site")
start_crawl(url="https://example.com", max_pages=500)
get_crawl_status(crawl_id="abc123")   # poll until complete
create_sitemap(name="Site Map", job_id="abc123")
export_sitemap(sitemap_id=1, format="json")

SEO Improvement Pipeline

run_seo_audit(crawl_id="abc123")
get_seo_issues(crawl_id="abc123", severity="error")
generate_suggestions(crawl_id="abc123", focus_areas=["seo"])
apply_suggestion(crawl_id="abc123", suggestion_id="sug_1")
get_seo_score(crawl_id="abc123")

Automated Monitoring

create_schedule(name="Weekly", url="https://example.com", frequency="weekly")
create_webhook(name="Slack", url="https://hooks.slack.com/...", events=["crawl.completed", "changes.detected"])
compare_crawls(crawl_id_old="abc123", crawl_id_new="def456")

Custom Data Extraction

test_extraction_rule(url="https://example.com", rule_type="css", selector="h1")
create_extraction_rule(name="Page Title", rule_type="css", selector="h1", target="text")
create_extraction_rule(name="Price", rule_type="css", selector=".price", target="text", match_all=True)
start_crawl(url="https://example.com", extraction_rule_ids="1,2")
get_extracted_data(crawl_id="abc123")

Site Migration — Redirect Map

start_crawl(url="https://example.com", max_pages=1000)
create_sitemap(name="Migration Plan", job_id="abc123")
# Restructure sitemap, mark pages as redirect with destination URLs
export_sitemap(sitemap_id=1, format="redirects", redirect_code=301)
# Or get structured JSON:
export_sitemap(sitemap_id=1, format="redirects_json")

WordPress Integration

Use IATO together with WordPress.com's MCP server to crawl, audit, and fix your WordPress site — all from Claude Desktop.

Setup

Add both connectors in Claude Desktop (Settings → Connectors → Add):

How It Works

IATO MCP (Analyze)          Claude (Orchestrate)          WordPress MCP (Edit)
       |                           |                              |
  Crawl site ──────────────> Identify issues ──────────────> Fix titles
  SEO audit ───────────────> Prioritize fixes ─────────────> Update meta
  Content gaps ────────────> Generate content ─────────────> Edit posts
  Broken links ────────────> Map to pages ─────────────────> Fix links

WordPress-Ready Tools

IATO includes 3 tools that produce WordPress-ready output with post slugs:

Example Prompts

Claude chains tools from both servers automatically. Just describe what you want and it picks the right tools.

Configuration