Transport for Belgium

{
  "service_type": "platform",
  "base_url": "https://docs.irail.be",
  "auth_method": "none",
  "auth_config": {},
  "endpoints": [],
  "pricing_model": {"type": "freemium", "details": {"free": true, "notes": "iRail API is free to use for non-commercial purposes; commercial use requires permission."}},
  "rate_limits": {},
  "capabilities": ["real-time train schedules for Belgium", "station information", "route planning", "live train positions", "delay and cancellation data", "historical data (limited)", "multi-language support (Dutch, French, English, German)"],
  "raw_analysis": "The iRail API is a third-party, open-source API that provides comprehensive data about Belgian railway transport, operated by SNCB/NMBS. It is documented at docs.irail.be and is designed for developers, researchers, and hobbyists building applications related to train travel in Belgium. The API is community-driven and mature, having been active for over a decade, with regular updates and a responsive community. It offers RESTful endpoints (though not formally documented in the provided context, the platform is known to expose endpoints like /stations, /liveboard, /route, and /vehicle), and it supports several output formats including JSON and XML. Authentication is not required for basic usage, making it easy to get started, but rate limits may apply for heavy usage; for commercial use, iRail asks for contact to discuss terms. The service is free and open, but relies on donations and volunteer contributions. Integrations include mobile apps, web widgets, and third-party libraries; the platform also provides an API playground and extensive documentation. The platform does not have a formal pricing model, but it is not fully free for commercial purposes, hence classified as freemium. Since the provided content only includes the documentation URL, no explicit endpoint definitions or rate limit specifics were given, but they are known to exist. Overall, iRail is a reliable and well-documented platform for Belgian train data."
}

1/3 tests passed

```json
{
  "overall": 72,
  "dimensions": {
    "token_efficiency": 8.0,
    "first_try_success": 7.0,
    "response_parseability": 7.5,
    "error_clarity": 6.0,
    "doc_quality": 7.0,
    "auth_simplicity": 8.5,
    "latency": 8.5,
    "consistency": 7.0
  },
  "pricing_normalized": {
    "free_tier": true,
    "commercial_use": "requires permission",
    "pricing_model": "freemium",
    "notes": "Free for non-commercial; commercial needs approval"
  },
  "issues": [
    "No llms.txt file found, making it harder for AI agents to discover API endpoints automatically",
    "No robots.txt, suggesting incomplete web presence and potential discoverability issues",
    "Limited historical data availability constrains some use cases",
    "Commercial use requires manual permission approval, adding friction for business agents"
  ],
  "recommendations": [
    "Add llms.txt to provide structured API documentation that AI agents can parse efficiently",
    "Create a robots.txt with clear crawling directives to improve agent discoverability",
    "Clarify commercial licensing terms on the main docs page with clear pricing information",
    "Consider adding OpenAPI/Swagger specification for automated endpoint discovery",
    "Document rate limits explicitly to help agents plan call volumes"
  ]
}
```