Skip to content
Back to skills

Code Standards

ASecurity

Extract and generate coding best practices from PR review comments. Use when the user asks to "extract best practices", "analyze PR comments", "generate coding standards", "create best practices from PRs", or "update coding guidelines from reviews".

  • 666 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added February 8, 2026
documentationtypescriptgobashtestingcode-reviewgitdocumentation

Works with

  • cli

Security analysis

A100/100

Scanned February 10, 2026

npx -y skills add majiayu000/claude-skill-registry --skill code-standards --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Code Standards?

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

Security grade badge for Code Standards
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-code-standards/badge)](https://www.skillsdirectory.com/skills/majiayu000-code-standards)

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
---
name: best-practices-extractor
description: Extract and generate coding best practices from PR review comments. Use when the user asks to "extract best practices", "analyze PR comments", "generate coding standards", "create best practices from PRs", or "update coding guidelines from reviews".
version: 1.0.0
---

# Best Practices Extractor

Extract coding best practices from PR review comments and codebase analysis to generate `.claude/best-practices/` documentation.

## When to Use

- Initial setup of coding standards for a project
- Regular updates to capture new team feedback
- Analyzing PR comments to document patterns
- Generating best practices from code review history

## Extraction Methods

All outputs are saved to `.claude/pr-review-comments/`.

### Method 1: Incremental Update (Recommended)

For regular updates - fast and tracks state:

```bash
cd plugin/skills/best-practices-extractor
bash scripts/incremental_update.sh OWNER REPO_NAME
```

First run does full extraction; subsequent runs fetch only new PRs.

**Output**: `.claude/pr-review-comments/{repo}_inline_comments.ndjson`

### Method 2: Full Extraction

For first-time setup or comprehensive analysis:

```bash
bash scripts/extract_pr_comments.sh REPO_NAME
```

**Output**: `.claude/pr-review-comments/{repo}_inline_comments.ndjson`

### Method 3: Sort by File Tree

Organize extracted comments by directory structure:

```bash
bash scripts/sort_comments_by_filetree.sh .claude/pr-review-comments/{repo}_inline_comments.ndjson
```

**Output**: `.claude/pr-review-comments/sorted/`

## Generating Best Practices

After extracting comments:

1. **Identify patterns** - Look for recurring feedback themes
2. **Categorize** - Group by type (naming, error-handling, type-safety, etc.)
3. **Create documentation** - Generate `.claude/best-practices/` files

### Output Structure

```
.claude/best-practices/
├── README.md              # Index
├── naming-conventions.md
├── error-handling.md
├── type-safety.md
├── testing.md
└── ...
```

### Guideline Format

Each guideline should include:

```markdown
## [Number]. [Title]

**Guideline:** [Clear statement]

**Why:** [Impact explanation]

**Example:**
```typescript
// Good
{example}

// Bad
{counter_example}
```

**Source:** PR #{number}
```

## Prerequisites

- **GitHub CLI (`gh`)**: Authenticated - check with `gh auth status`
- **jq**: JSON processor - check with `jq --version`

## Integration

This skill **generates** best practices. The `code-reviewer` agent and `/4_review` command **validate** against them.

Workflow:
1. Extract PR comments (this skill)
2. Analyze and generate `.claude/best-practices/`
3. Reviews automatically validate against best practices

## Additional Resources

- **`references/default-categories.md`** - Example category definitions
- **`references/scripts-guide.md`** - Detailed script documentation

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.