Indexing Guard for AI agents

Let your agent check pages before they ship.

Coding agents can generate thousands of pages in an afternoon. Connect Indexing Guard and the same agent can check them: what goes in the sitemap, what was held back, and what to fix in the data.

It works with Claude Code, Cursor, Codex and any client that speaks MCP. Use the API key you already have. The free trial works too: up to 100 URLs per check, 3 checks a day, with counts only. The page-by-page reasons come with a paid plan.

Connect

Store your key as an environment variable first, for example INDEXING_GUARD_KEY. Don't paste it into a chat.

Claude Code

claude mcp add --transport http indexing-guard https://guard.myartmyrules.com/mcp \
  --header "Authorization: Bearer $INDEXING_GUARD_KEY"

Cursor

In .cursor/mcp.json:

{
  "mcpServers": {
    "indexing-guard": {
      "url": "https://guard.myartmyrules.com/mcp",
      "headers": { "Authorization": "Bearer ${env:INDEXING_GUARD_KEY}" }
    }
  }
}

Codex

In ~/.codex/config.toml:

[mcp_servers.indexing-guard]
url = "https://guard.myartmyrules.com/mcp"
bearer_token_env_var = "INDEXING_GUARD_KEY"

Any other MCP client

URL:     https://guard.myartmyrules.com/mcp
Header:  Authorization: Bearer <your key>
Transport: Streamable HTTP

Step-by-step pages: Claude Code · Cursor · OpenAI Codex · Gemini CLI.

What your agent can do

• check_pages: send a build's pages and get the counts, every page held back with its reason, page warnings, and the input to fix for each reason. It can also return the sitemap files.
• list_runs: the last 50 checks with their counts, to compare one build with the next.

Things to ask

• "Check the pages in dist/routes.json with Indexing Guard and tell me what was held back."
• "Which pages are missing a source? Add the source links from our data and check again."
• "How did this week's builds compare with last week's?"

What your agent can't do

• Change your account. No tool rotates keys, changes plans or touches billing. Those stay with you.
• Go past your plan. Every check counts against the same limits as your build, and an inactive plan gets an error.
• See anything but your results. Agents get the same answers your build gets: counts, reasons and fixes. They see nothing from other customers, and we store counts, not your URLs.

Your build stays in charge

Agents are for checking and fixing. For deploys, keep the build step from the docs, which falls back to your own sitemap if we can't be reached.

Search engines decide what they index and rank. Indexing Guard controls what you submit.

Questions: terrell@myartmyrules.com