Agent-Readiness Audit for Public URLs

Choose this endpoint when you need a deterministic, reproducible, evidence-backed audit of a public URL's readiness for AI agent consumption. Unlike LLM-based site evaluations, this endpoint produces the same score for the same input every time, with published weights and a recompute recipe — ideal for compliance checks, marketplace listing qualification, CI/CD pipeline integration, or benchmarking multiple sites objectively. Prefer it over manual review or generic crawlers when you need structured, weighted per-check evidence rather than a narrative summary.

• URL is not publicly reachable — returns error or low score with failed checks
• Invalid or malformed URL input — schema validation error
• Site blocks automated audit requests (e.g. aggressive bot protection) — checks may fail with no evidence
• Transient network errors reaching the target site — partial or incomplete breakdown returned
• Mismatched input schema fields (e.g. wrong types for score or max_score) — request rejected