GovTribe

Update memories MCP tool

Updates an existing durable user memory after locating the correct entry; prefer updating canonical memories over creating duplicates. Can revise memory text, reinforce or weaken memories, or set a future expires_at review date. Legacy memories without a future review date must include expires_at before they can be updated.

Examples

Update the canonical memory instead of creating a duplicate when the user changes a durable preference. Use lifecycle arguments for memories that should remain reviewable by GovTribe AI. Legacy or permanent memories that do not already have a future review date must include a future expires_at before they can be updated.

At least one update action is required: memory, expires_at, strength, or reinforce. To reinforce or weaken an existing lifecycle-managed memory without changing the text, pass reinforce or strength and leave memory out.

Update an existing memory about saved-search cadence

{
  "tool": "Search_Memories",
  "arguments": {
    "query": "saved search notification cadence",
    "search_mode": "keyword",
    "size": 5
  }
}
{
  "tool": "Update_Memories",
  "arguments": {
    "memory_id": "<memory_govtribe_id>",
    "memory": "The user prefers weekly saved-search notification examples by default, but uses instant notifications for time-sensitive recompetes."
  }
}

Reinforce a lifecycle-managed memory

{
  "tool": "Update_Memories",
  "arguments": {
    "memory_id": "<memory_govtribe_id>",
    "reinforce": true,
    "reason": "The same preference was useful in another saved-search workflow."
  }
}

Weaken a lifecycle-managed memory and set a new review date

{
  "tool": "Update_Memories",
  "arguments": {
    "memory_id": "<memory_govtribe_id>",
    "expires_at": "2026-10-15",
    "strength": "weak",
    "reason": "The preference may only apply during the current capture planning period."
  }
}