AgentPilot — API AI-Readiness & MCP Server Generator

What This Actor Does

Audits public OpenAPI specifications for AI suitability and automatically generates run-ready typescript/python Model Context Protocol (MCP) server code packages.

Who It Is For

• AI developers
• software engineers
• CTOs
• product managers
• system architects

Why It Is Useful

Instead of being a generic API caller, this Actor evaluates schemas against strict LLM usability standards (description completeness, schema types, auth complexity) and generates fully functional stdio MCP servers on the fly.

Input Fields

• startUrls: public website URLs to process.
• maxItems: maximum URLs to process.
• includeAiAnalysis: optional AI enrichment when API keys are configured.
• industryHint: optional category hint.

Example Input

{
  "startUrls": [
    {
      "url": "https://example.com"
    }
  ],
  "maxItems": 1,
  "includeAiAnalysis": false,
  "industryHint": "local business"
}

Example Output

{
  "inputUrl": "https://example.com",
  "status": "success",
  "score": 64,
  "opportunityScore": 45,
  "findings": {},
  "missingItems": [],
  "recommendations": [],
  "createdAt": "2026-08-13T04:40:52.965Z"
}

Output Schema

The Actor declares .actor/output_schema.json and links the results output to the default dataset items URL. This helps Apify Console, API consumers, and AI agents discover where run results are stored.

Limitations

Public HTML only. No login-protected scraping, private APIs, legal compliance claims, or browser UI testing.

Local Development

npm install
npm run build
npm test
npm run lint
apify run

Deployment

$apify push