motion-review
Skill1 starsChanged 3 months agoMITView on GitHub
motion-pro-max 独立动效审查器 — 5 维打分 + 三级质量门。Blocking>0 或总分<80 不能交付。每次 motion-pro-max 生成动效代码后必须调用。
3 other projects use a version of this.
Adopted by
Most recently taken up by
- PicsArt/gen-ai-skillstheir own version, first seen Nov 8, 2026
- whoamnesia/astudio-motiontheir own version, first seen Nov 8, 2026
- lymanballif/motion-reviewtheir own version, first seen Oct 25, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- PicsArt/gen-ai-skills★ 6same name
- whoamnesia/astudio-motion★ 1same name
- lymanballif/motion-review★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repositoryOne other file it gives its agents.
- motion-pro-maxSkill
The file
Files
Motion Review — 动效交付审查器
独立审查动效代码。调用 reviewer subagent(opencode 的只读审查 agent)进行 5 维打分 + 三级质量门审查。因为是独立 agent,审查上下文与生成上下文隔离,降低自我审查的确认偏误。
Blocking > 0 或总分 < 80 = 不能交付。
审查工作流
motion-pro-max 生成代码
↓
/review(你)
↓
Phase 1: 读取规则 → 构建审查 prompt
Phase 2: 调用 reviewer subagent(独立上下文,无偏误)
Phase 3: 收结果 → 5 维打分 → 修复 Blocking → 输出报告Phase 1: 构建审查 Prompt
读取以下规则文件,提取核心审查标准(quality-gates.md 是 🔴/🟡/🔵 清单的唯一权威,以此为准):
../motion-pro-max/references/quality-gates.md— Blocking(5)/Warning(8)/Preference(9) 清单../motion-pro-max/references/motion-principles.md— 七原理评估标准../motion-pro-max/references/motion-system.md— 弹簧/时长合规标准../motion-pro-max/rubric.md— 5 维评分标准(权威)
Phase 2: 审查标准(传给 reviewer)
你是 motion-pro-max 的动效代码审查员。审查以下 GSAP 动画代码。
## 🚫/⚠️/💭 质量门
按 Phase 1 读取的 quality-gates.md 逐项检查:
- 🚫 Blocking 5 项:任一不通过 = 不能交付
- ⚠️ Warning 8 项:列出全部
- 💭 Preference 9 项:按上下文判断
对每一项必须回显:通过 / 不通过 + 证据位置 + 修复建议。
禁止只报计数——逐项列出,防止漏检。
## 🎯 七原理评估(逐项 0-4 分,满分 28)
- Momentum: 入场 ease-out / 退场 ease-in / 退<入 / 无 ease-in 入场
- Hierarchy: L1→L5 运动顺序 / 最重先动 / 无同时动
- Weight: 不同元素不同 duration / badge轻 hero重 / 3+质量级
- Delay: stagger 指数递减 / 总揭示≤600ms / 超10个用batch
- Follow-through: 附属元素多动 50-150ms / 阴影滞后主体
- Anticipation: 反向预备+主动作 / back.out(1.2-1.7) / 非滥用
- Overshoot: 分级使用 / Toast中过冲 / Modal微过冲 / 不滥用
## ⚙️ 弹簧/时长合规
- 时长以 SKILL.md 26 场景速查表为精确真值;bouncy 未用于 Modal/表单/L3 按钮
- 未使用 ease(CSS默认)/ linear
- 退场时长 < 入场时长
- Stagger 总揭示 ≤ 600ms
- 五级运动层级匹配(L3 150-300ms / L4 300-500ms 等)
## 输出格式
对每个问题给出: 严重级别 | 位置 | 描述 | 修复建议
最后输出:
- 🚫 Blocking 5 项逐项回显:通过 / 不通过 + 证据 + 修复建议(禁止只报计数)
- ⚠️ Warning 8 项逐项回显:有 / 无
- 每个原理评分 (0-4) + 理由(7项×4=28满分)
- 弹簧合规通过数/总数(5项)
审查代码:
{粘贴 motion-pro-max 生成的代码}Phase 3: 5 维评分计算
收到 reviewer 审查结果后,按 ../motion-pro-max/rubric.md(权威评分标准)计算 5 维评分:
1. 设计系统一致性 (20%) = (通过的 Blocking 数 / 5) × 20
2. 动效物理感 (20%) = 七原理总分 / 28 × 20
3. 性能 (20%) = compositor-only(7) + quickTo(5) + will-change(5) + batch(3)
4. 可访问性 (20%) = reduced-motion(8) + focus-visible(6) + contrast(6)
5. 框架正确性 (20%) = React cleanup(8) + scope(6) + registerPlugin(6)
总分 = 五项之和(上限 100,下限 0)交付规则
Blocking > 0 → 不能交付。列出所有 Blocking,逐项自动修复,最多 3 轮。
Score < 80 → 不能交付。逐维度说明扣分原因和修复方向。
Warning > 3 → 建议修。列出所有 Warning,用户确认。
Score >= 80 → 通过。输出 Review Report。评级
| 总分 | 等级 | 处理 |
| 90-100 | 🟢 卓越 | 直接交付 |
| 80-89 | 🟢 优秀 | 修复 Warning 后交付 |
| 60-79 | 🟡 合格 | 必须修复 Warning |
| 40-59 | 🟠 需改进 | 至少修复 3 项原理问题 |
| < 40 | 🔴 不合格 | 打回重做 |
报告模板
═══════════════════════════════════════
Motion Pro Max 动效审查报告
(由 reviewer subagent 独立审查)
═══════════════════════════════════════
## Quality Gates(逐项回显,见附录)
🚫 Blocking: X/5 — [未通过项列表]
⚠️ Warning: N/8 — [列表]
💭 Preference: N/9
## 5-Dimension Scores
| 维度 | 得分 | 扣分原因 |
|------|------|---------|
| 1. 设计系统一致性 | /20 | |
| 2. 动效物理感 | /20 | |
| 3. 性能 | /20 | |
| 4. 可访问性 | /20 | |
| 5. 框架正确性 | /20 | |
| **总分** | **/100** | |
## Verdict
[PASS / FAIL] — [说明]
## Fixes Applied
[自动修复的变更列表]
## Recommendations
[给用户的建议]
## 附录: Blocking/Warning 逐项回显清单
| # | 级别 | 检查项 | 结果 | 证据位置 | 修复建议 |
|---|------|--------|------|---------|---------|
| 1 | 🚫 | outline: none | 通过/不通过 | ... | ... |
| ... | | (5 Blocking + 8 Warning 全列) | | | |铁律
- 必须用 reviewer subagent(opencode 只读审查 agent),不能自己在主会话里审
- Blocking 不通过 = 自动修复后重审,最多 3 轮
- 原理评分必须给出具体理由,不能只说"不错"
- Blocking/Warning 必须逐项回显结果,禁止只报计数
- 不修改业务逻辑代码,只审动效部分
- 首次用先读规则文件(quality-gates.md 清单如有更新,以它为准)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

