Integration surfaces for the Open Synaps API
All endpoints are proxied through /api/agi/* and forward to the Synaps AGI cognitive engine. Multi-session support via X-Session header.
Authentication
API Token
Set AUTH_TOKEN env var on the engine. Pass via Authorization: Bearer <token> header.
Multi-Tenant
Use X-Tenant header for tenant isolation. Each tenant gets independent memory and state.
/chatCoreSend a message to the AGI cognitive pipeline. Supports multi-session, multi-tenant, and mode selection (plan, reason, default).
curl -X POST https://opensynapse.in/api/agi/chat \
-H "Content-Type: application/json" \
-d '{"message": "Explain quantum computing"}'/statusMonitoringRetrieve engine status including inference count, active experts, memory entries, and system health.
curl https://opensynapse.in/api/agi/status
/healthzHealthLightweight health check endpoint. Returns 200 if the engine is running.
curl https://opensynapse.in/api/agi/healthz
/tracesReasoningRetrieve reasoning traces from recent cognitive pipeline executions. Includes mode, experts used, and confidence scores.
curl https://opensynapse.in/api/agi/traces
/self/stateSelf-AnalysisGet the self-analysis state: novelty, drift, forgetting metrics. Powers belief monitoring and Theory of Mind.
curl https://opensynapse.in/api/agi/self/state
/drift/checkSafetyRun drift detection across model parameters. Returns drift score and per-dimension analysis.
curl https://opensynapse.in/api/agi/drift/check
/api/causalWorld ModelRetrieve the causal graph structure. Returns nodes (causes) and edges (effects) with strength values.
curl https://opensynapse.in/api/agi/api/causal
/metricsMetricsPrometheus-compatible metrics endpoint for monitoring inference latency, memory usage, and throughput.
curl https://opensynapse.in/api/agi/metrics
/memory/search?q=...MemorySearch across memory systems (Short-term, Long-term, Permanent) using keyword matching. Returns ranked results with provenance.
curl "https://opensynapse.in/api/agi/memory/search?q=quantum"