mcpField
D
daily-reward-and-streak-review

daily-reward-and-streak-review

github.com/sylphxaiMarketing & growth152 installsUpdated today
Repository

About daily-reward-and-streak-review

A review skill for product designers shaping daily rewards and streak-based retention. It produces a compact loop design with state models, reward tables, metrics, and safeguards for trust and economy health.

Key features

  1. 01Identify core return behavior
  2. 02Review cadence and streak states
  3. 03Assess recovery and notifications
  4. 04Check economy and abuse risks
  5. 05Create state models, tables, and metrics

Use cases

  1. 01Design a daily reward loop
  2. 02Review missed-streak handling
  3. 03Assess paid repairs and VIP boosts
  4. 04Set guardrails for retention experiments

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/sylphxai/skills --skill daily-reward-and-streak-review

After installing, type/daily-reward-and-streak-reviewin Claude Code to confirm it loaded.

Bundled filesagents/openai.yamlreferences/daily-reward-streak-systems.md

Details

From the project README

Daily Reward And Streak Review

Use this skill to make daily retention loops motivating, fair, and economy-safe.

Workflow

  1. Identify the core behavior the product wants to reinforce and the user value of returning.
  2. Read references/daily-reward-streak-systems.md.
  3. Map reward cadence, streak state, missed-day handling, notifications, economy impact, and abuse controls.
  4. Separate healthy habit reinforcement from punishment, pressure, or low-value farming.
  5. Treat paid repair, VIP multipliers, gacha currency, comeback bonuses, and notification copy as trust-sensitive monetization surfaces.
  6. Produce a complete but compact loop design with state model, reward table, metrics, and experiment guardrails. Prefer short bullets over long prose; never omit metrics.

When not to use

  • Do not use for generic advice the base model already handles without this skill's specific artifact contract.

Guardrails

  • Do not make the product depend on anxiety or loss aversion as the main value.
  • Do not create rewards that undermine pricing, progression, or marketplace fairness.
  • Provide recovery paths for missed days when long-term trust matters more than strict streak purity.
  • Do not make paid repair the only practical recovery path or advertise it with panic, countdown, or shame copy.
  • Do not let VIP multipliers or gacha-currency ladders become pay-to-win, inflationary, or unfair to non-paying players.

Output format

Core return behavior:
Reward promise:

Loop design:
- cadence, trigger, reward, recovery, cap

State model:
- <state> -> entry, exit, user message

Metrics:
- activation, repeat, long-term retention, fatigue, economy impact, trust/support, abuse false positives

Economy/support risks:
- <risk> -> mitigation

Trust and monetization pressure:
- paid repair:
- VIP/gacha impact:
- notification copy:
- player support path:

Comments

Markdown supported