{
  "name": "EvalKit",
  "description": "Score AI outputs on configurable quality criteria using LLM-as-judge evaluation. A single API call to evaluate accuracy, relevance, safety, and more.",
  "url": "https://evalkit.dev",
  "documentationUrl": "https://evalkit.dev/docs",
  "openApiUrl": "https://evalkit.dev/openapi.json",
  "llmsTxtUrl": "https://evalkit.dev/llms.txt",
  "authentication": {
    "type": "bearer",
    "description": "API key from your EvalKit dashboard. Format: ek_live_... or ek_test_..."
  },
  "capabilities": [
    "Evaluate LLM output quality on built-in criteria (accuracy, relevance, coherence, safety, tone, completeness, conciseness, groundedness)",
    "Define custom evaluation criteria",
    "Batch evaluate up to 20 outputs in a single request",
    "Get structured scores (0-1) with per-criterion reasoning",
    "Detect issues and receive improvement suggestions"
  ],
  "contactEmail": "support@evalkit.dev"
}