Stores a learning, pattern, or decision in an agent memory system for future recall.
- What it does
- Records a learning, pattern, or decision in a memory system so it can be recalled later. It stores the entry in PostgreSQL with BGE embeddings, auto-detects the learning type when none is given, extracts tags from the content, and returns a confirmation with an ID. Supported types include working solution, architectural decision, codebase pattern, failed approach, and error fix.
- When to use it
- Use it when something learned during work should be kept for later retrieval, such as a fix that worked, a design decision, a recurring codebase pattern, or an approach that failed. It is invoked manually with a short description of the learning.
- Requirements
- Requires a PostgreSQL database with BGE embeddings, the CLAUDE_OPC_DIR environment variable, the uv runtime, and the store_learning.py script in the referenced repository. The skill itself ships no scripts; it only provides instructions.
Remember - Store Learning in Memory
Store a learning, pattern, or decision in the memory system for future recall.
Usage
Or with explicit type:
Examples
What It Does
- Stores the learning in PostgreSQL with BGE embeddings
- Auto-detects learning type if not specified
- Extracts tags from content
- Returns confirmation with ID
Learning Types
Execution
When this skill is invoked, run:
Auto-Type Detection
If no --type specified, infer from content:
- Contains "error", "fix", "bug" → ERROR_FIX
- Contains "decided", "chose", "architecture" → ARCHITECTURAL_DECISION
- Contains "pattern", "always", "convention" → CODEBASE_PATTERN
- Contains "failed", "didn't work", "don't" → FAILED_APPROACH
- Default → WORKING_SOLUTION
