Open Targets Platform REST API

{
  "service_type": "platform",
  "base_url": "https://api.opentargets.org",
  "auth_method": "none",
  "auth_config": {},
  "endpoints": [],
  "pricing_model": {
    "type": "free",
    "details": {
      "note": "Open Targets Platform is a public, open-access resource funded by a consortium of pharmaceutical companies and research institutions (GSK, Sanofi, Bayer, EMBL-EBI, etc.). Data and API access are free."
    }
  },
  "rate_limits": {},
  "capabilities": [
    "programmatic retrieval of target-disease association data",
    "evidence-based target identification and prioritisation",
    "disease and phenotype data access",
    "drug and clinical candidate information",
    "target annotation and tractability information",
    "GraphQL and REST interface access",
    "batch queries and data download for large-scale analysis",
    "integration with Ensembl, ChEMBL, UniProt, Europe PMC and other bioinformatics resources"
  ],
  "raw_analysis": "Open Targets Platform is a comprehensive, publicly funded bioinformatics platform that aggregates and scores evidence linking therapeutic targets to human diseases. It is primarily used by drug discovery researchers, computational biologists, and translational scientists in pharma and academia.\n\nThe REST API provides programmatic access to the Platform's data, enabling users to query target-disease associations, evidence types (genetic, somatic, expression, literature, animal models, pathways), drugs, and disease ontologies. This makes it suitable for building automated pipelines for target identification, prioritisation, and hypothesis generation.\n\nThe service is mature and well-maintained, backed by a consortium including EMBL-EBI, GSK, Sanofi, Bayer and others. It has an active release cycle and stable endpoints (v2 REST API and a GraphQL API). No authentication is required for standard public use, and rate limiting (if any) is generally permissive for reasonable research use.\n\nThe platform also offers bulk data downloads (Parquet, JSON) for large-scale analysis, plus a web interface, documentation, and community support via GitHub and a discourse forum. It is considered a cornerstone resource in computational drug discovery alongside similar efforts like Pharos and Open Targets Genetics."
}

0/3 tests passed

{
  "overall": 62,
  "dimensions": {
    "token_efficiency": 8.0,
    "first_try_success": 5.0,
    "response_parseability": 9.0,
    "error_clarity": 6.0,
    "doc_quality": 7.5,
    "auth_simplicity": 9.0,
    "latency": 4.0,
    "consistency": 5.0
  },
  "pricing_normalized": {
    "model": "free",
    "notes": "Open-access resource funded by pharma consortium (GSK, Sanofi, Bayer, EMBL-EBI). No paywall, no API-key cost, no usage tiers."
  },
  "issues": [
    "All three verification checks failed with DNS resolution errors ('Name or service not known') — domain could not be resolved at test time, indicating either a transient network/DNS failure in the test harness or an outage/unreachable endpoint.",
    "Latency measurements from failed checks (60-266ms) reflect DNS failure timing, not real endpoint responsiveness, so actual latency is unknown.",
    "No security headers or robots.txt/llms.txt could be validated, so agent-discoverability artifacts are unconfirmed.",
    "No observed uptime data means consistency score is based on reputation (a well-established EMBL-EBI-hosted resource) rather than measured stability.",
    "GraphQL schema is extensive; without an llms.txt or LLM-oriented doc, agents may burn tokens discovering query shapes."
  ],
  "recommendations": [
    "Re-run checks against the canonical host (platform.opentargets.org) and its API subdomain to confirm reachability before relying on this platform.",
    "Verify GraphQL endpoint availability (api.platform.opentargets.org/api/v4/graphql) directly, as this is the primary agent-facing interface.",
    "Publish an llms.txt and concise GraphQL schema summary to reduce agent token spend on query discovery.",
    "Document rate limits, max page sizes, and batch/download best practices explicitly for agent consumers.",
    "Expose a status page / uptime history to improve consistency scoring for agent recommenders.",
    "Report actual API latency (p95) for representative queries to inform agent planning/retry logic."
  ]
}