Skip to content

Connect Sitepager to Your AI Assistant

Sitepager MCP connects your AI assistant directly to your scan data. Your assistant can run scans, analyze results, propose fixes, and re-run to confirm everything is resolved, all in one session.

What used to be a manual loop across multiple tools becomes a single conversation with your assistant.

Every issue your assistant surfaces comes from an actual Sitepager scan. It reads structured data from the API. It does not crawl or guess.


Once connected, your assistant can:

  • Create and run scans and wait for results
  • Summarize issues across your entire site
  • Identify broken pages and broken links
  • Detect pages added or removed since the baseline
  • Review visual changes against the baseline
  • Investigate SEO issues on specific pages
  • Retrieve Lighthouse reports from a scan and review performance issues


  1. Log in to Sitepager
  2. Go to Settings > API Keys
  3. Create a new key and copy it
Sitepager Settings showing API Keys section
Create an API key in Settings > API Keys.

Your key starts with sp_live_. Keep it secure. It gives access to all scans and results in your workspace.


The Sitepager MCP server is distributed as an npm package, @sitepager/mcp-server, and runs via npx. Add it to your AI assistant’s MCP configuration using the steps below.

Add to your .mcp.json file:

{
"mcpServers": {
"sitepager": {
"command": "npx",
"args": ["-y", "@sitepager/mcp-server"],
"env": {
"SITEPAGER_API_KEY": "paste-your-api-key-here"
}
}
}
}

Replace paste-your-api-key-here with the key you copied from the previous step.

Add to your MCP settings using the same command and SITEPAGER_API_KEY environment variable.

Any MCP-compatible assistant works. Use npx -y @sitepager/mcp-server as the command and set SITEPAGER_API_KEY as the environment variable.


Start a new session and ask:

Prompt
List my Sitepager scans.

Your assistant should return a list of your scans. If it does, the MCP connection is working.


Run a Sitepager scan for https://example.com and summarize the results.
Run the [scan name] scan and wait for it to complete.
Summarize the highest-severity issues in the latest run of [scan name].
Are there any broken pages or broken links in the latest run of [scan name]?
Which pages have visual differences from the baseline in [scan name]?
What SEO issues exist on /pricing in the latest run of [scan name]?
Review the Lighthouse results for the landing page from the latest run of [scan name]. Summarize the scores and flag anything that needs attention.