From helix
Show learning system health - insight counts, tag distribution, effectiveness, feedback loop status.
How this skill is triggered — by the user, by Claude, or both
Slash command
/helix:helix-statsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
```bash
HELIX="$(cat .helix/plugin_root)"
python3 "$HELIX/lib/memory/core.py" health
Returns: status (HEALTHY/NEEDS_ATTENTION), total_insights, total_edges, connected_ratio, avg_edges_per_insight, by_tag, effectiveness, with_feedback, recent_feedback, loop_coverage, causal_ratio, issues.
For distribution analysis and tuning constant calibration:
python3 "$HELIX/lib/memory/core.py" stats
Returns JSON with:
effectiveness — histogram of insight effectiveness (10 buckets)context_spread — distribution of generality scores (5 buckets)velocity — count by recent_uses valuetop_velocity — most actively used insightstop_connected — highest graph degree insightssession_log — outcome counts by type and agentnpx claudepluginhub enzokro/crinzo-plugins --plugin helixObservability dashboard that analyzes session stats, agent tracking, hook performance, and learning health. Provides actionable recommendations for skill calibration.
Analyzes Claude Code session data to produce a usage analytics dashboard with epistemic protocol coverage, friction analysis, and growth timeline.
Analyzes memory system state: health diagnostics, topic coverage scoring, knowledge gap detection, and filesystem validation. Use before new project phases or after backfill.