用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/InugamiDev/ultrathink-oss --skill code-review命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | code-review |
| description | Multi-pass code review covering logic correctness, security, performance, style, and maintainability |
| layer | hub |
| category | workflow |
| triggers | ["/code-review","review this code","check this code","review my changes","is this code good"] |
| inputs | [{"target":"File(s), diff, PR, or code snippet to review"},{"focus":"Specific concern area (optional -- security, performance, etc.)"},{"context":"What the code is supposed to do (optional but helpful)"}] |
| outputs | [{"reviewReport":"Structured report with findings across all passes"},{"findings":"Categorized issues (critical, major, minor, nit)"},{"suggestions":"Concrete improvement recommendations with code examples"}] |
| linksTo | ["fix","refactor","test","scout","optimize"] |
| linkedFrom | ["cook","team","ship","fix","refactor"] |
| preferredNextSkills | ["fix","refactor","test"] |
| fallbackSkills | ["scout","debug"] |
| riskLevel | low |
| memoryReadPolicy | selective |
| memoryWritePolicy | selective |
| sideEffects | ["Reads source files","Produces review report in working memory","Does NOT modify any files"] |
Perform thorough, multi-pass code review that catches bugs, security vulnerabilities, performance issues, and maintainability problems before they reach production. Each pass focuses on a different dimension of code quality.
A code review is not a checklist exercise. It is an act of empathy -- reading code as a future maintainer who needs to understand it, modify it, and trust it.
Focus: Does the code do what it is supposed to do?
any types?Focus: Can this code be exploited?
Focus: Will this code perform well at scale?
Focus: Will a future developer understand and safely modify this code?
Focus: Does this change fit well in the broader system?
Categorize findings by severity:
Produce the review report using the template below.
# Code Review Report
## Summary
**Files reviewed**: [count]
**Overall assessment**: Approve | Request Changes | Needs Discussion
**Critical issues**: [count]
**Major issues**: [count]
---
## Critical Issues
### [C1] [Title]
**File**: [path:line]
**Category**: Logic | Security | Performance
**Description**: [what is wrong]
**Impact**: [what could happen]
**Suggestion**:
```[language]
// suggested fix
...
...
## Usage
### Review specific files
/code-review src/lib/auth.ts src/app/api/login/route.ts
### Review with focus
/code-review Security review of the payment processing module
### Review recent changes
/code-review Review the changes in the last commit
### Review a PR
/code-review Review PR #42
## Examples
### Example: Finding a logic bug
**Code**:
```typescript
function getDiscount(items: Item[]): number {
if (items.length > 10) return 0.15;
if (items.length > 5) return 0.10;
if (items.length > 0) return 0.05;
}
Finding (Critical): Function has no return statement when items.length === 0. Returns undefined, which will cause NaN when used in arithmetic.
Code:
const query = `SELECT * FROM users WHERE email = '${email}'`;
Finding (Critical): SQL injection vulnerability. User-provided email is interpolated directly into the query string. Use parameterized queries instead.
Code:
const results = users.map(user => {
const orders = await db.orders.findMany({ where: { userId: user.id } });
return { ...user, orders };
});
Finding (Major): N+1 query problem. For 1000 users, this makes 1001 database queries. Use a single query with a JOIN or batch the lookups.
fix or refactor skill implements changes.Unified design foundations — design system architecture, tokens, component specs, visual principles, creative vision, figma integration, plus brand design system loader (66 real brands via DESIGN.md). Absorbs design, design-system, design-systems, design-principles, design-router, creative-vision, figma, design-md.
Render, summarize, and present markdown documents and structured content in multiple output modes
Ultra UI skill - combines Google's DESIGN.md spec (machine-readable design tokens) with the ui-ux-pro-max knowledge base (91 styles, 161 palettes, 73 font pairings, 161 products, 104 UX guidelines, 25 chart types). Generates lint-clean DESIGN.md files, validates token references and WCAG contrast, exports Tailwind/DTCG tokens, and diffs design systems version-over-version.
基于 SOC 职业分类
