mneme_prime
Builds a preflight context bundle for a new session by combining recent session notes with topic-relevant matches within a set token budget.
Instructions
Preflight context bundle for a new session. Combines recent session-typed docs and topic-relevant matches inside a token budget. v1.0 uses the full injection format; Phase F.5 adds the keypoints/ref Adaptive Context Layer. Pass session_id (from CLAUDE_SESSION_ID) to activate per-session injection deduplication and progressive format selection (full → keypoints → ref as context fills).
Tool call template
Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {
"task_description": "<task_description>"
},
"name": "mneme_prime"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| session_id | No | Caller session identifier (e.g. CLAUDE_SESSION_ID). Enables per-session injection deduplication and progressive format selection (full to keypoints to ref). | |
| budget_tokens | No | ||
| topic_doc_count | No | ||
| task_description | Yes | ||
| recent_session_count | No |