Community ops weekly review
Skill ShiroRyu9/community-ops-kit/skills/community-ops-weekly-review
Use when producing community weekly reviews, weekly reports, metric summaries, decision notes, retrospective plans, or next-week operating priorities. Connects activity data, onboarding, creator signals, paid/member care, support risks, content assets, and owner decisions into a practical review.From its SKILL.md
npx -y skills add ShiroRyu9/community-ops-kit --skill community-ops-weekly-reviewAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 23 days oldThe repository was created 23 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
3.4 KB, 639 tokens by cl100k_base, as published. Nobody here has run it
Community Ops Weekly Review
Use this skill to turn a week of community work into decisions and next actions.
Project Inputs
Get the review period and timezone, durable source records, action states, metric sources and denominators, user or creator signals, known risks, data gaps, and owner decisions from the user or approved project sources.
The skill supplies the review lens and output structure. It does not include the community's weekly actions, counts, revenue, retention, member records, or conclusions. If records are incomplete, review the available evidence and name the missing sources; never fill gaps with repository examples or assumed trends.
Review Lens
Review the week by operating loop, not by raw message volume:
entry -> first action -> participation -> submission/proof -> feedback/showcase -> retention/upgrade -> review
Inputs To Check
- activity launches, reminders, submissions, votes, winners, showcases;
- newcomer path and first-action signals;
- creator/high-quality content signals;
- paid member or high-value user follow-ups;
- support, payment, access, safety, or moderation risks;
- product/website/account conversion signals if available;
- unverified metrics and data gaps.
- daily operating records, execution receipts, verification results, and material state changes.
Prefer durable operating records over report summaries when the two conflict. Daily reports and dashboards are derived views; they should not become the only evidence for a weekly conclusion.
Output Pattern
Localize the visible labels to the user's requested language.
**Weekly Judgment**
[What improved, broke, or needs a decision.]
**Review Scope**
- Period/timezone: ...
- Data sources and denominator: ...
**Key Actions**
| Area | Prepared | Executed/verified | Evidence | Decision needed |
|---|---|---|---|---|
**User-Path Review**
entry -> first action -> participation -> proof -> feedback/showcase -> next action
**Opportunities**
- [Concrete improvement with surface and owner.]
**Risks / Gaps**
- [Missing data, approval, support issue, rule gap, or unclear benefit.]
**Next-Week Priorities**
1. [P0 + owner + due date]
2. [P1 + owner + due date]
3. [P2 + owner + due date]
Approval Boundary
Weekly review outputs are recommendations and drafts. Require explicit confirmation before sending reminders, contacting or mentioning users, publishing showcases or winner lists, settling rewards, changing roles/permissions, or making official metric and product claims.
Quality Bar
Do not turn the review into a generic strategy essay. Each conclusion needs evidence: count, link, artifact, visible action, user segment, or an explicit data gap.
Keep prepared, pending approval, executed, verified, blocked, ambiguous, and cancelled distinct. An automation run or approved draft does not prove that a public action happened. Feed verified decisions and material state changes back into the durable operating record for the next cycle.
Avoid unsupported revenue or retention claims. Say "possible signal" when the evidence is partial.
What ships with it: 1 file
279 B alongside SKILL.md
agents/
- openai.yaml279 B