meridianlearnlisted

# /meridian:learn — Execution Learning System Capture patterns and mistakes as persistent rules injected into future subagent prompts. ## Arguments - `<rule>` — The learning to capture (freeform text) - `--list` — Show all learnings - `--prune` — Remove stale/unused learnings - `--delete <id>` — Remove a specific learning - `--scope <global|project|phase>` — Scope for the learning (default: project) - `--source <manual|execution|review|debug>` — How the learning was discovered (default: manual) - `--extract [<phase-dir>]` — Extract structured learnings from phase artifacts - `--extract --all` — Extract from all completed phases missing LEARNINGS.md - `--extract --pending` — List phases that need extraction (dry-run, no writes) ## Keywords learn, learning, rule, pattern, remember, lesson, mistake, capture, inject, extract ## Procedure ### Step 1: Route by argument **If `--list`**: Show all learnings: ```bash PYTHONPATH=$MERIDIAN_HOME uv run --project $MERIDIAN_HOME -- python -c " import json from scripts.db import open_project from scripts.learnings import list_learnings with open_project('.') as conn: learnings = list_learnings(conn) print(json.dumps(learnings, indent=2, default=str)) " ``` Display grouped by scope (global → project → phase), showing ID, rule, source, category, and applied_count. Also show extraction status: ```bash PYTHONPATH=$MERIDIAN_HOME uv run --project $MERIDIAN_HOME -- python -c " import json from pathlib import Path from scripts.extrac