recalllisted

## Intent Execute **recall** to retrieve relevant prior decisions, rationale, and patterns from memory to maintain consistency and learn from past interactions. **Success criteria:** - Relevant memories are retrieved based on query - Prior decisions and their rationale are surfaced - Consistency with past reasoning is maintained - Retrieved information is grounded with timestamps and context **Compatible schemas:** - `schemas/output_schema.yaml` ## Inputs | Parameter | Required | Type | Description | |-----------|----------|------|-------------| | `query` | Yes | string | What to recall: topic, decision type, or pattern name | | `memory_scope` | No | enum | Where to search: `session` (current), `project` (CLAUDE.md), `global` (all). Default: `project` | | `time_range` | No | object | Filter by time: `{ after: "2024-01-01", before: "2024-01-31" }` | | `include_rationale` | No | boolean | Whether to include decision rationale. Default: true | | `similarity_threshold` | No | number | Minimum relevance score (0.0-1.0). Default: 0.5 | ## Procedure 1) **Parse query intent**: Understand what is being recalled - Identify if query is about a decision, pattern, fact, or context - Extract key terms for memory search - Determine if exact match or semantic similarity needed 2) **Scope memory search**: Identify which memory stores to query - `session`: Current conversation context - `project`: CLAUDE.md, knowledge files, local docs - `global`: All accessible memor