haabe

Use to assess regulatory applicability for products that may fall under AI regulation (EU AI Act, Article 50 transparency).

A/B test CLAUDE.md instruction changes against eval benchmarks. Capture baselines, test variants, compare results.

Lint canvas files for staleness, missing fields, inconsistent evidence types, and orphaned references. Run periodically or before major transitions.

Record findings from completed offline human tasks (interviews, observations, outreach) back into the canvas. The re-entry point after /mycelium:handoff.

Evaluate Mycelium's own process effectiveness. Measures cycle velocity, discard trends, confidence calibration, gate effectiveness, regression rate. Run quarterly or every 20 cycles.

Generate structured handoff materials for offline human tasks (interviews, observations, outreach). Creates actionable briefs with phone-friendly capture templates.

Guide for conducting Torres-style story-based user interviews with bias mitigation and JTBD lens.

Use to evaluate whether current work aligns with Better Value Sooner Safer Happier. Run at diamond completion and periodically.

Test fixture skill that writes to canvas AND carries the Preflight block. Check 31 must pass.

Parallel agent orchestration for OST exploration. Fan-out multiple solution explorations, fan-in results to compare and select winners.

Update canvas sections with new evidence. Ensures canvas stays current as the single source of truth.

Use to build or update an Opportunity Solution Tree from research data. Never from brainstorming.