extracting-code-structure

This collection includes 10 specialized skills covering file discovery, code analysis, text processing, and data extraction:

Core Skills

• [ast-grep](./skills/ast-grep/) - Structural code search and refactoring
• [bat](./skills/bat/) - Enhanced file viewer with syntax highlighting
• [code-structure](./skills/code-structure/) - Extract code structure efficiently
• [fd](./skills/fd/) - Fast file discovery with parallel search
• [fzf](./skills/fzf/) - Interactive fuzzy finder

Data Processing Skills

• [jq](./skills/jq/) - JSON data extraction and manipulation
• [yq](./skills/yq/) - YAML data extraction and manipulation

Text Processing Skills

• [ripgrep](./skills/ripgrep/) - Fast text search with one-shot patterns
• [sd](./skills/sd/) - Intuitive find & replace with JavaScript regex

Analysis Skills

• [tokei](./skills/tokei/) - Code statistics analysis by language

macOS (using Homebrew)

```bash

# Install all tools with one command

brew install ast-grep bat fd fzf jq ripgrep sd tokei yq

```

Install from source (Cargo)

```bash

# Install all Rust-based tools

cargo install ast-grep fd-find ripgrep sd tokei

cargo install --locked bat # bat requires --locked flag

```

Manual Installation Links

• fzf: https://github.com/junegunn/fzf#installation
• bat: https://github.com/sharkdp/bat#installation
• ripgrep: https://github.com/BurntSushi/ripgrep#installation
• jq: https://stedolan.github.io/jq/download/
• yq: https://github.com/mikefarah/yq#install

Each skill contains a SKILL.md file with:

• When to use the tool
• Common workflows and patterns
• Integration examples with other tools
• Progressive disclosure to reference documentation

Common Workflows

```bash

# Interactive file search and edit

fd -e py | fzf --preview="bat --color=always {}" | xargs vim

# Code analysis pipeline

code-structure → ripgrep → ast-grep → bat

# Configuration management

fd → jq/yq → fzf → sd

# Statistics-driven refactoring

tokei → fd → ripgrep → ast-grep → tokei

```

To add a new command-line tool to this collection:

1. Create Documentation

Add a simplified documentation file to the docs/ directory:

• Name it after the tool (e.g., newtool.md)
• Include only core features, installation methods, and basic usage
• Keep it concise and focused on practical examples

2. Generate SKILL File

Use AI to generate a SKILL.md file following the project's specification:

```bash

# Create the skill directory

mkdir -p skills/newtool

# Ask AI to generate SKILL.md based on:

# 1. The docs/newtool.md content

# 2. The existing skills/*SKILL.md patterns

# 3. This project's philosophy of context efficiency

```

3. Update README

Add the new skill to the appropriate category in README.md:

• Include installation commands
• Add it to relevant workflows

4. Test Integration

Ensure the new tool integrates well with existing tools in the collection.

Copy these skills to Claude Code for enhanced functionality:

Claude API

• Upload skills as a zip file
• Specify skills in the container parameter

Claude Desktop/claude.ai

• Navigate to ~/.claude/skills/ or .claude/skills/
• Copy skill directories directly

Droid

• Navigate to ~/.factory/skills/ or .factory/skills/
• Copy skill directories directly

Skill Requirements

Based on [skills.md](./skills.md), each skill needs:

#### Required Structure

```yaml

---

name: your-skill-name # lowercase/numbers/hyphens only

description: What the skill does and when to use it # third person, specific

---

```

#### Best Practices

• Use gerund form naming: processing-files, analyzing-code
• Keep SKILL.md under 500 lines
• Split content into separate files for complex skills
• Include both what skill does AND when to use it
• Follow progressive disclosure patterns

#### Example Skill Structure

```

skills/tool-name/

├── SKILL.md (main instructions)

├── scripts/ (utility scripts)

├── docs/ (reference documentation)

└── examples/ (input/output examples)

```

Integration Patterns

• Tool composition: Chain multiple tools for powerful workflows
• Context efficiency: Extract only what you need
• Progressive disclosure: Load details when needed
• Error handling: Validate inputs and provide helpful error messages

These skills prioritize:

• Context efficiency - Extract only what you need
• Tool composition - Combine tools for powerful workflows
• Performance - Use optimized, fast tools
• Interactivity - Leverage fuzzy finders and previews
• Progressive disclosure - Details when needed, not before

MIT License

See the [LICENSE](./LICENSE) file for the full license text.