Skip to content
Back to skills

Roll Review

ASecurity

Self code review step in the TCR workflow. Runs after each micro-step is completed and before commit, checking code quality, security, and design issues.

  • 4 stars
  • 0 votes
  • 0 copies
  • 5 views
  • Added September 12, 2026
code-qualitybashsqlgitsecurityperformance

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 12, 2026

npx -y skills add aibot88/sec_skill_store --skill roll-review --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Roll Review?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Roll Review
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/aibot88-roll-review/badge)](https://www.skillsdirectory.com/skills/aibot88-roll-review)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
hidden: true
name: roll-.review
license: MIT
model: sonnet
allowed-tools: "Read, Bash(git:*)"
description: Self code review step in the TCR workflow. Runs after each micro-step is completed and before commit, checking code quality, security, and design issues.
---

# WK Self Code Review

**Quality gate for the TCR loop** - Self-review after each micro-step is completed and before commit.

## Position in WK Workflow

```
TCR Loop:
  Write Test → Run Test → ✅ Green → Self Review → Commit
                                          ↓
                                     Critical?
                                   Yes → Fix → Redo
                                    No → Proceed
```

## When Triggered

- **Auto-triggered**: After each TCR micro-step in `$roll-build` / `$roll-fix`
- **Manual trigger**: When the user wants to review current changes

## When Not to Use

- Docs-only changes (README, CHANGELOG) with no code diff
- Test pyramid / coverage standards check (use `$roll-.qa`)
- Post-deploy production sampling (use `$roll-sentinel`)

## Review Scope

```bash
# Default: review staged changes (recommended for TCR)
$roll-.review staged

# Review all uncommitted changes
$roll-.review unstaged

# Review specific files
$roll-.review files src/utils.ts
```

## Review Dimensions (6 Core Dimensions)

```
┌─────────────────────────────────────────────────────────┐
│  WK Quality Checklist                                   │
├─────────────────────────────────────────────────────────┤
│  ✅ Correctness     - Logic is correct, no bugs         │
│  ✅ Security        - No vulnerabilities, input valid.  │
│  ✅ Maintainability - Clear naming, sound structure     │
│  ✅ Performance     - No performance pitfalls           │
│  ✅ Testability     - Easy to test, edge cases covered  │
│  ✅ Scope           - Focused on current task, no       │
│                       unrelated changes                 │
└─────────────────────────────────────────────────────────┘
```

## Severity Levels and Decisions

| Level | Definition | Decision |
|-------|-----------|----------|
| 🔴 **Critical** | Bug, security vulnerability | **Must fix**, redo TCR |
| 🟡 **Warning** | Maintainability issue | **Recommend fix** or document |
| 🟢 **Suggestion** | Minor optimization | Optional, proceed with commit |
| ✅ **Pass** | No issues | Proceed with commit |

## Output Format

```markdown
## Self Review Report
**Scope**: staged (2 files, +45/-12 lines)

### 🔴 Critical (Must Fix)
| File | Line | Issue | Action |
|------|------|-------|--------|
| auth.ts | 23 | SQL injection | Use parameterized query |

### 🟡 Warnings
- utils.ts:45 - Magic number, consider: `const MAX_RETRY = 3`

### ✅ Passed
- Naming conventions
- Error handling
```

## TCR Integration

In each micro-step of `$roll-build`:

```markdown
**Micro-Step X: [Description]**

1. Write/Update Test
2. Run Test → ✅ Green
3. **$roll-.review staged**
   - 🔴 Critical? → Fix → Redo step
   - 🟡 Warning? → Quick fix or document
   - ✅ Pass? → Proceed
4. git commit -m "tcr: description"
```

## WK Principle Alignment

- **Agent-First**: Structured review checklist, executable by AI
- **Check Phase**: Local quality control
- **Micro-steps**: Small fast steps, each review < 100 lines
- **TCR**: Can only commit after passing self-check, ensuring repo quality

Files in this skill

  • SKILL.md3.7 KB
  • source.json658 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.