Skip to main content

​
Overview

Marks a session as completed in the LongMem database. This allows the system to properly track session boundaries and improve memory recall.

​
Authentication

Requires Bearer token if authToken is configured in daemon settings. 
Authorization: Bearer <your-token>

​
Request

​
session_id
string
required
Session identifier to mark as completed

​
Response

​
status
string
Always “ok” on success

​
Example

cURL
curl -X POST http://localhost:38741/session/end \
  -H "Content-Type: application/json" \
  -d '{
    "session_id": "my-coding-session-123"
  }'
Response
{
  "status": "ok"
}

​
Use Cases

​
Clean Session Boundaries

Call this when ending a coding session to mark the completion timestamp: 
await fetch('http://localhost:38741/session/end', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    session_id: currentSessionId
  })
});

​
Integration Cleanup

Most integrations (OpenCode plugin, Claude Code hooks) call this automatically when the AI session ends or the tool exits.

​
Implementation Details

From daemon/routes.ts:245-256:
  • Looks up the session in the cache or database
  • Marks the session as completed with a timestamp
  • Removes the session from the in-memory cache
  • No-op if session doesn’t exist (returns “ok” anyway)

​
Behavior Notes

  • Idempotent: Safe to call multiple times with the same session_id
  • Optional: Sessions don’t require explicit ending — they’ll be marked completed on daemon restart
  • No data loss: Observations and prompts are already saved when this is called

​
Error Responses

​
Missing session_id

{
  "error": "session_id required"
}
Status: 400

Build docs developers (and LLMs) love

Get started for freeTalk to us