Skip to main content
Record a significant event, interaction, or conversation as a chronological log entry. Episodes provide an audit trail separate from memories. Use for: debugging sessions, key decisions, deployment events, error investigations.

​
Parameters

​
content
string
required
Episode content. Be descriptive: include what happened, why, and the outcome. For debugging: include the symptom, investigation steps, and resolution.
​
contentType
string
default:"text"
Content format.
ValueDescription
textFree-form text (default)
jsonStructured JSON data
messageConversation or chat message format
​
source
string
Origin of the episode. Examples: "session:abc123", "file:src/auth.ts", "debug:OOM-investigation".
​
refs
object[]
References linking this episode to knowledge graph entities/edges. Enables traversal from episodes to related concepts.Each ref has:
  • refType: "node" to reference an entity, "edge" to reference a relationship
  • refId: ID of the referenced entity or edge
​
scope
string
default:"project"
Storage scope.
ValueDescription
projectStored under the current project (default)
globalShared across all projects

​
Example

{
  "content": "Debugged OOM error in GraphService. Root cause: unbounded connection pool. Fixed by adding limit of 10.",
  "contentType": "text",
  "source": "debug:OOM-investigation",
  "refs": [
    { "refType": "node", "refId": "entity_graphservice_id" }
  ]
}