Skip to main content

​
Base URL

https://your-instance.minns.ai
All endpoints are prefixed with /api/. There is no /v1/ prefix.
The base URL defaults to https://your-instance.minns.ai for local development. Update this to your production URL when deploying.

​
Request format

  • All request bodies use JSON (Content-Type: application/json).
  • All responses return JSON.
  • Timestamps are u64 values representing nanoseconds since Unix epoch.
  • IDs are unsigned integers: u64 for most identifiers, u128 for EventId.

​
Server-generated fields

The server automatically generates certain fields if you omit them:
FieldAuto-generatedDescription
event.id✅ YesUUID generated if omitted
event.timestamp✅ YesCurrent timestamp generated if omitted
context.fingerprint✅ YesHash computed if set to 0 or omitted
Best practice: Omit id, timestamp, and fingerprint and let the server generate them. This ensures consistency and reduces client complexity.

​
Null vs. omitted fields

ScenarioExampleBehavior
Required"agent_id": 1Must be present with a valid value
Server-generatedOmit "id"Server auto-generates
Optional with defaultOmit "limit"Uses default value (usually 10)
Optional, nullable"deadline": nullCan be omitted or explicitly set to null
Optional, non-nullable"causality_chain": []Can be omitted (defaults to []) but cannot be null

​
Default query limits

ParameterDefaultMaximum
limit101000
Claims per event—10
Claim confidence threshold—0.7 minimum for storage

​
System profiles

FeatureFreeNormal
Max graph size50,000 nodes1,000,000 nodes
Memory cache1,000 items10,000 items
Strategy cache500 items5,000 items
Redb cache64 MB256 MB
Louvain analyticsDisabledEnabled

​
Endpoint groups

Events

Ingest events, retrieve event history, and view episodes.

Memory

Query long-term memories by agent or by context.

Strategies

Retrieve and search learned behavioral strategies.

Claims

List, get, and search extracted facts.

Graph

Explore the knowledge graph, analytics, and communities.

System

Health checks and server status.