Skill Claude CodeCodex
Reads data/skill-telemetry/invocations.jsonl and produces daily stats — dead skills (zero invocations in last 30d), hot skills (>10 invocations/day on average), failure rates per skill, median wallclock per skill. Auto-publishes the dead-skill list to data/skill-telemetry/dead-skills-pending-review.md so future…
Skill Claude CodeCodex
Daily 04:00 CT housekeeping over data/approval-gates/. Expires pending past expiresat, archives gates >7 days old. Telegram only on change.
Skill Claude CodeCodex
Sundays 18:00 CDT. Composes a public state-of-the-company narrative covering the past 7 days — what shipped, what was caught, what's being tested next. Reads private inputs (open handoffs by specialist, latest sunday-review, recent decisions, recent P-gate catches) and emits a sanitized public-safe markdown narrative…
Skill Claude CodeCodex
Score a piece of writing against VOICE.md's rubric. Returns score + findings. ADVISORY ONLY — not a hard gate. Author (Cooper or LLM drafter) reviews findings and decides whether each is a real issue or false positive. The hard gate for public output is public-content-scrubber (financials).
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: