MCP HubMCP Hub
返回技能列表

repomix

Elios-FPT
更新于 Today
26 次查看
1
在 GitHub 上查看
aidesign

关于

Repomix is a tool that packages entire code repositories into single, AI-friendly files. It is primarily used to generate context for LLMs like Claude, enabling AI analysis, codebase snapshots, and security audits. This makes it ideal for preparing a codebase for an AI to review or summarize.

技能文档

Repomix Skill

Repomix is a powerful tool that packs entire repositories into single, AI-friendly files. Perfect for when you need to feed codebases to Large Language Models (LLMs) or other AI tools like Claude, ChatGPT, and Gemini.

When to Use This Skill

Use this skill when:

  • User needs to package a codebase for AI analysis
  • Preparing repository context for LLM consumption
  • Generating codebase snapshots for documentation
  • Analyzing third-party libraries or repositories
  • Creating AI-friendly representations of code projects
  • Investigating bugs across large codebases
  • Performing security audits on repositories
  • Generating context for implementation planning

Core Capabilities

1. Repository Packaging

Repomix packages entire repositories into single files with:

  • AI-optimized formatting with clear separators
  • Multiple output formats (XML, Markdown, JSON, Plain text)
  • Git-aware processing (respects .gitignore)
  • Token counting for LLM context management
  • Security checks for sensitive information

2. Remote Repository Support

Can process remote repositories without cloning:

  • Shorthand: npx repomix --remote yamadashy/repomix
  • Full URL: npx repomix --remote https://github.com/owner/repo
  • Specific commits: npx repomix --remote https://github.com/owner/repo/commit/hash

3. Comment Removal

Strips comments from supported languages when needed:

  • Supported: HTML, CSS, JavaScript, TypeScript, Vue, Svelte, Python, PHP, Ruby, C, C#, Java, Go, Rust, Swift, Kotlin, Dart, Shell, YAML
  • Enable with: --remove-comments or config file

Installation

Check if installed first:

repomix --version

Install using preferred method:

# npm
npm install -g repomix

# yarn
yarn global add repomix

# bun
bun add -g repomix

# Homebrew (macOS/Linux)
brew install repomix

Basic Usage

Package Current Directory

# Basic packaging (generates repomix-output.xml)
repomix

# Specify output format
repomix --style markdown
repomix --style json
repomix --style plain

# Custom output path
repomix -o custom-output.xml

Package Specific Directory

repomix /path/to/directory

Package Remote Repository

# Shorthand format
npx repomix --remote owner/repo

# Full URL
npx repomix --remote https://github.com/owner/repo

# Specific commit
npx repomix --remote https://github.com/owner/repo/commit/abc123

Command Line Options

File Selection

# Include specific patterns
repomix --include "src/**/*.ts,*.md"

# Ignore additional patterns
repomix -i "tests/**,*.test.js"

# Disable .gitignore rules
repomix --no-gitignore

# Disable default ignore patterns
repomix --no-default-patterns

Output Configuration

# Output format
repomix --style markdown  # or xml, json, plain

# Output file path
repomix -o output.md

# Remove comments
repomix --remove-comments

# Show line numbers
repomix --no-line-numbers  # disable line numbers

Security & Analysis

# Run security checks
repomix --no-security-check  # disable security scanning

# Copy to clipboard
repomix --copy  # copy output to clipboard

# Verbose output
repomix --verbose

Configuration

# Use custom config file
repomix -c custom-config.json

# Initialize new config
repomix --init  # creates repomix.config.json

Configuration File

Create repomix.config.json in project root:

{
  "output": {
    "filePath": "repomix-output.xml",
    "style": "xml",
    "removeComments": false,
    "showLineNumbers": true,
    "copyToClipboard": false
  },
  "include": ["**/*"],
  "ignore": {
    "useGitignore": true,
    "useDefaultPatterns": true,
    "customPatterns": [
      "additional-folder",
      "**/*.log",
      "**/tmp/**"
    ]
  },
  "security": {
    "enableSecurityCheck": true
  }
}

Ignore Patterns

.repomixignore File

Create .repomixignore for Repomix-specific ignore patterns (same format as .gitignore):

# Build artifacts
dist/
build/
*.min.js

# Test files
**/*.test.ts
**/*.spec.ts
coverage/

# Large files
*.mp4
*.zip

# Sensitive files
.env*
secrets/

Precedence Order

  1. CLI ignore patterns (-i flag)
  2. .repomixignore file
  3. Custom patterns in config file
  4. .gitignore file (if enabled)
  5. Default patterns (if enabled)

Output Formats

XML Format (Default)

Best for structured AI consumption:

repomix --style xml

Markdown Format

Human-readable with syntax highlighting:

repomix --style markdown

JSON Format

For programmatic processing:

repomix --style json

Plain Text

Simple concatenation:

repomix --style plain

Use Cases & Examples

1. Code Review Preparation

# Package feature branch for AI review
repomix --include "src/**/*.ts" --remove-comments -o feature-review.md --style markdown

2. Security Audit

# Package third-party library for analysis
npx repomix --remote vendor/library --style xml -o audit.xml

3. Documentation Generation

# Package with docs and code
repomix --include "src/**,docs/**,*.md" --style markdown -o context.md

4. Bug Investigation

# Package specific modules
repomix --include "src/auth/**,src/api/**" -o debug-context.xml

5. Implementation Planning

# Full codebase context for planning
repomix --remove-comments --copy

Token Management

Repomix automatically counts tokens for:

  • Individual files
  • Total repository
  • Per-format output

Use token counts to manage LLM context limits:

  • Claude: ~200K tokens
  • GPT-4: ~128K tokens
  • GPT-3.5: ~16K tokens

Security Considerations

Sensitive Data Detection

Repomix uses Secretlint to detect:

  • API keys and tokens
  • Passwords and credentials
  • Private keys
  • AWS secrets
  • Database connection strings

Disable if needed:

repomix --no-security-check

Best Practices

  1. Always review output before sharing
  2. Use .repomixignore for sensitive files
  3. Enable security checks for unknown codebases
  4. Avoid packaging .env files
  5. Check for hardcoded credentials

Performance Optimization

Large Repositories

Repomix uses worker threads for parallel processing:

  • Efficiently handles large codebases
  • Example: facebook/react processed 29x faster (123s → 4s)

Optimization Tips

# Exclude unnecessary files
repomix -i "node_modules/**,dist/**,*.min.js"

# Process specific directories only
repomix --include "src/**/*.ts"

# Disable line numbers for smaller output
repomix --no-line-numbers

Workflow Integration

With Claude Code

# Package and analyze in one workflow
repomix --style markdown --copy
# Then paste into Claude for analysis

With CI/CD

# Generate codebase snapshot for releases
repomix --style markdown -o release-snapshot.md

With Git Hooks

# Pre-commit hook to generate context
repomix --include "src/**" -o .context/latest.xml

Common Patterns

Full Repository Package

repomix --remove-comments --style markdown -o full-repo.md

Source Code Only

repomix --include "src/**/*.{ts,tsx,js,jsx}" -i "**/*.test.*"

Documentation Bundle

repomix --include "**/*.md,docs/**" --style markdown

TypeScript Project

repomix --include "**/*.ts,**/*.tsx" --remove-comments --no-line-numbers

Remote Analysis

npx repomix --remote owner/repo --style xml -o analysis.xml

Troubleshooting

Issue: Output Too Large

# Exclude unnecessary files
repomix -i "node_modules/**,dist/**,coverage/**"

# Process specific directories
repomix --include "src/**"

Issue: Missing Files

# Disable .gitignore rules
repomix --no-gitignore

# Check ignore patterns
cat .repomixignore

Issue: Sensitive Data Warnings

# Review flagged files
# Add to .repomixignore
# Or disable checks: --no-security-check

Implementation Workflow

When user requests repository packaging:

  1. Assess Requirements

    • Identify target repository (local/remote)
    • Determine output format needed
    • Check for sensitive data concerns
  2. Configure Filters

    • Set include patterns for relevant files
    • Add ignore patterns for unnecessary files
    • Enable/disable comment removal
  3. Execute Packaging

    • Run repomix with appropriate options
    • Monitor token counts
    • Verify security checks
  4. Validate Output

    • Review generated file
    • Confirm no sensitive data
    • Check token limits for target LLM
  5. Deliver Context

    • Provide packaged file to user
    • Include token count summary
    • Note any warnings or issues

Best Practices

  1. Start with defaults: Run basic repomix first, then refine
  2. Use .repomixignore: Better than CLI flags for complex patterns
  3. Enable security checks: Especially for third-party code
  4. Choose right format: XML for AI, Markdown for humans
  5. Monitor token counts: Stay within LLM limits
  6. Remove comments: When focusing on logic over documentation
  7. Version output: Include in .gitignore, regenerate as needed
  8. Test patterns: Verify include/exclude work as expected

Related Tools

  • Context7: For up-to-date library documentation
  • Git: For repository history analysis
  • Secretlint: For security scanning
  • Token counters: For LLM context management

Additional Resources

快速安装

/plugin add https://github.com/Elios-FPT/EliosCodePracticeService/tree/main/repomix

在 Claude Code 中复制并粘贴此命令以安装该技能

GitHub 仓库

Elios-FPT/EliosCodePracticeService
路径: .claude/skills/repomix

相关推荐技能

llamaguard

其他

LlamaGuard是Meta推出的7-8B参数内容审核模型,专门用于过滤LLM的输入和输出内容。它能检测六大安全风险类别(暴力/仇恨、性内容、武器、违禁品、自残、犯罪计划),准确率达94-95%。开发者可通过HuggingFace、vLLM或Sagemaker快速部署,并能与NeMo Guardrails集成实现自动化安全防护。

查看技能

sglang

SGLang是一个专为LLM设计的高性能推理框架,特别适用于需要结构化输出的场景。它通过RadixAttention前缀缓存技术,在处理JSON、正则表达式、工具调用等具有重复前缀的复杂工作流时,能实现极速生成。如果你正在构建智能体或多轮对话系统,并追求远超vLLM的推理性能,SGLang是理想选择。

查看技能

evaluating-llms-harness

测试

该Skill通过60+个学术基准测试(如MMLU、GSM8K等)评估大语言模型质量,适用于模型对比、学术研究及训练进度追踪。它支持HuggingFace、vLLM和API接口,被EleutherAI等行业领先机构广泛采用。开发者可通过简单命令行快速对模型进行多任务批量评估。

查看技能

langchain

LangChain是一个用于构建LLM应用程序的框架,支持智能体、链和RAG应用开发。它提供多模型提供商支持、500+工具集成、记忆管理和向量检索等核心功能。开发者可用它快速构建聊天机器人、问答系统和自主代理,适用于从原型验证到生产部署的全流程。

查看技能