Skill fixes from @rudycelekli. A reviewer recomputed every calculation, and the guidance stays firm where a change would otherwise have hedged it.
- Math:
- offers (1.0.1 → 1.0.2): multiply sequential lifts rather than adding them (#809).
- churn-prevention (2.0.2 → 2.0.3): dunning ROI stated net of tooling, at 5–14x (#810). Health scores put every signal on one 0–100 scale and handle missing signals explicitly (#829).
- co-marketing (2.0.2 → 2.0.3): marketplace fees are recurring costs, not CAC, and Arrows is listed on the HubSpot Marketplace (#812).
- referrals (2.2.0 → 2.2.1): program return measured after serving costs (#820).
- ab-testing (2.0.1 → 2.0.2): test duration accounts for the number of arms and exposure (#819).
- Methods:
- attribution (1.1.4 → 1.1.5): intent-to-treat analysis vs. on-off tests, and an interval that includes zero is not a win (#815).
- revops (2.2.0 → 2.2.1): calibrate scoring on snapshots taken at decision time and on holdouts (#817).
- pricing (2.1.3 → 2.1.4): the direction of the Van Westendorp "too cheap" curves (#821).
- onboarding (2.0.1 → 2.0.2): early activation measured separately from mature retention (#826). Card-required trials judged on visitor→paid, with ChartMogul's ~30% trial→paid (over 5x no-card) as context (#827).
- Platform facts:
Skill ZIP assets
Complete source catalog: 50 skills.
Claude web target: 50 skills with reviewed short descriptions.
Choose <name>-claude-web.zip for the documented single-folder web upload format.
Source ZIPs retain original metadata; they are not substitutes for omitted web ZIPs.
Live client upload, account limits, and invocation behavior have not been verified.
Web omissions (also recorded in claude-web-manifest.json):
- None.