OBSERVABILITY API
Integrate governed reliability analysis, correction, and audit into any AI pipeline with a single authenticated API call.
Governed by Default
Every request is audit-logged and governed end-to-end.
Fast Integration
Single API call. No additional infrastructure required.
Benchmark-Validated
All corrections validated against governance benchmarks.
Real-Time Events
Subscribe to governance events via webhook or event stream.
ENDPOINTS
Everything you need to govern AI output reliability, from analysis through audit.
/analyzeAnalyze AI Output
CoreSubmit any AI-generated output for comprehensive reliability analysis. Detects contradictions, hallucinations, confidence drift, and semantic inconsistencies.
/correctApply Governed Correction
GovernanceRequest a governed correction for a flagged output. Preview with dry_run before committing. All corrections are benchmark-validated before delivery.
/audit/{reference_id}Retrieve Audit Summary
AuditRetrieve the full governance audit record for any prior analysis. Includes reliability status, dimensions scored, and governance verdict.
/healthCheck System Health
StatusReturns current platform governance status, uptime, and reliability metrics. Use for integration health checks and monitoring.
INTEGRATION EXAMPLE
POST /analyze
Authorization: Bearer oap_••••••••
Content-Type: application/json
{
"model_id": "gpt-4o",
"output": "...",
"context": {}
}{
"reference_id": "ref_••••••••",
"governance_verdict": "approved",
"reliability_status": "validated",
"dimensions": {
"consistency": "pass",
"grounding": "verified",
"confidence": "calibrated",
"coherence": "pass"
},
"audit_logged": true
}AUTHENTICATION
Authenticate with your API key
All requests are user-scoped. Every request is audit-logged. Get your API key from the API Keys settings panel after signing in.