Engagement retention advisor
Skill ajitta/Game-Engagement-Retention-Skills/skills/engagement-retention-advisor
Produce integrated proposals that jointly improve moment-level engagement AND lifecycle retention for games, interactive apps, AI companions, fortune/reading apps, visual novels, learning apps, communities. Use when the request contains TWO deliverables — (a) in-session reward/moment design AND (b) lifecycle/return strategy — OR the moment-to-retention link is the question ('what in-session moment raises retention?'), including a moment complaint paired with a churn/return problem with no explicit deliverable ('combat feels flat and players churn') — OR the user asks to compare or prioritize a moment-level vs lifecycle-level investment ('better combat feel or a re-engagement push first?'). A retention metric cited only as motivation or success criterion is NOT a second ask ('D7 is low, fix the reward staging' → interaction-reward-moments). Retention-only → retention-strategy-designer. Outputs 3-5 integrated proposals pairing reward moments with retention mechanisms, plus metrics, experiments, domain ethics.From its SKILL.md
npx -y skills add ajitta/Game-Engagement-Retention-Skills --skill engagement-retention-advisorAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 2 stars2 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
6.8 KB, as published. Nobody here has run it
Engagement Retention Advisor
Use when the user requests both moment-level reward design and a retention strategy as one proposal set, when the moment-to-return link itself is the question, or when the user asks to compare or prioritize a moment-level vs a lifecycle-level investment. The value of integration is not concatenating two lists — it is connecting how a satisfying in-session moment becomes value for the next visit.
See references/domain-ethics.md for detailed per-domain ethical guardrails.
Routing
- User asks for both deliverables (moment design + retention/return strategy), or the moment-to-return link itself is the question → this skill. A moment complaint paired with a churn/return problem with no explicit deliverable ("combat feels flat and players churn") counts as the link question.
- User asks to compare, sequence, or prioritize a moment-level investment vs a lifecycle-level investment ("better combat feel or a re-engagement campaign first?") → this skill, in comparison mode (see Output).
- Retention mentioned only as a goal / metric / modifier ("D7 is low, I want to fix the staging", "satisfying enough that they return") → stay single-skill. If it is moment design →
interaction-reward-moments. - User asks for retention only →
retention-strategy-designer. - Ambiguous and the choice materially changes the output → ask one clarifying question.
- When routing away, invoke the target skill; if the Skill tool is unavailable, Read
../interaction-reward-moments/SKILL.mdor../retention-strategy-designer/SKILL.mdand follow its workflow.
Workflow
-
Build a product model:
- User intent, core value, core interaction, current loop, return event, natural usage cycle.
-
Split the problem into two layers:
- Moment layer: what is satisfying within a single session? (Use the ten lenses defined in
../interaction-reward-moments/SKILL.mdstep 2 — read that file rather than working from memory. If you need patterns, read../interaction-reward-moments/references/pattern-library.md.) - Lifecycle layer: why return tomorrow, next week, next month? (Use the
retention-strategy-designerdiagnosis: return event, usage cycle, lifecycle stage, leak points. If you need benchmarks / experiment design, read../retention-strategy-designer/references/retention-playbook.md.)
- Moment layer: what is satisfying within a single session? (Use the ten lenses defined in
-
Recombine only where the two layers reinforce each other:
- Every integrated proposal must contain both a reward moment and a retention loop.
- State the integration rationale for each: why this moment is not mere stimulation but leads to next-visit value.
- Moment is satisfying but has no future value → flag as "engagement only" and separate it.
- Loop drives return but has no in-session satisfaction → flag as "retention only" and redesign.
-
Per-domain ethics check (
references/domain-ethics.md):- Fortune / saju: no deterministic life prophecy, no fear framing, no paid emotional relief.
- AI companion / journaling: no dependency, no over-notification, no crisis-counselor role confusion, no misuse of sensitive data.
- Visual novel: no manipulative cliffhangers, no paywalling core endings.
- Learning apps: no guilt streaks; recovery-friendly progress.
- Games: no hidden odds or undisclosed loot-box probabilities on real-money rewards, no pay-to-avoid-frustration, no life-punishing FOMO events, no variable rewards chained to escalating ad consumption.
- Overall: optimize for user value, not opens or session length.
Output
Answer in the user's language (Korean input → Korean output); keep the section/table structure below — headings may stay in English or be translated.
Comparison mode: when the request is a which-first / prioritization question rather than a two-deliverable ask, skip the 3-5 integrated-proposal table. Output instead: a brief diagnosis of both layers, a sequencing recommendation with rationale (impact, dependency, measurement order), and at most one integrated proposal if the layers genuinely reinforce each other.
## Input interpretation
- Product / scene:
- Core value:
- Natural usage cycle:
- Recommended return event:
## 3-5 integrated proposals
| # | Proposal | In-session reward moment | Trigger / user action | Feedback / staging | Retention mechanism | Target segment / lifecycle stage | Return event / time window | Ethical guardrail |
|---|---|---|---|---|---|---|---|---|
(keep each cell within 2 sentences)
## Integration rationale
(1-2 lines per proposal: why this moment is not mere stimulation but leads to next-visit value)
## Measurement design
(per proposal: core metric + experiment method. Measure by cohort, watch the novelty effect for 2-4 weeks, track guardrail metrics — notification opt-out rate, uninstalls, complaints — in parallel)
## Items to view separately
- Strong reward moment only (engagement only):
- Strong retention only (retention only → needs redesign):
## Prioritization
(order by impact × implementation difficulty, with a 1-line rationale each)
Guardrails
Apply guardrails as design constraints before generating proposals — not as a post-hoc filter.
- Do not force-convert every good reward moment into a retention loop.
- Do not let a retention goal flatten a concrete scene design into generic lifecycle advice.
- Prefer transparent, user-benefiting loops over compulsion, FOMO, guilt, or anxiety.
- When a prohibited pattern is the request itself: state the refusal rationale explicitly in the body (do not bury it in a table-cell footnote) and redesign an ethical alternative that meets the same business goal. No quiet partial compliance.
- Per-domain prohibition list:
references/domain-ethics.md.