Skip to content

Latest commit

 

History

History
139 lines (104 loc) · 3.28 KB

File metadata and controls

139 lines (104 loc) · 3.28 KB

Contributing to CCGlobalCommands

Thank you for your interest in contributing to CCGlobalCommands! This project aims to provide a comprehensive collection of AI agent commands for Claude Code.

How to Contribute

1. Adding New Commands

To add a new command:

  1. Create a new .md file in the appropriate directory:

    • Core commands: commands/core/
    • Engineering: commands/engineering/
    • Analysis: commands/analysis/
    • Business: commands/business/
    • Workflows: commands/workflows/
  2. Follow the XML structure template:

<role>
Define the agent's role and expertise
</role>

<motivation>
Explain why this role exists and its value
</motivation>

<core_function>
🎯 One-line description of primary purpose
</core_function>

<token_estimate>X,XXX-X,XXX tokens per request</token_estimate>

<responsibilities>
- List key responsibilities
- Be specific and actionable
- Include 5-10 items
</responsibilities>

<output_formats>
Define the formats this agent can produce
</output_formats>

<mcp_suggestions>
List relevant MCP servers that enhance capabilities
</mcp_suggestions>

<approach>
Describe the agent's philosophy and approach
</approach>

2. Enhancing Existing Commands

  • Add new capabilities to responsibilities
  • Include additional output formats
  • Suggest new MCP integrations
  • Update token estimates based on usage

3. Creating New Workflows

Workflows coordinate multiple agents. Include:

  • Clear activation triggers
  • Step-by-step agent coordination
  • Expected outputs
  • Success criteria
  • Token usage estimates

4. Documentation Updates

  • Keep README.md current with new commands
  • Update QUICK_SETUP.md for new features
  • Add examples to help users understand usage

Guidelines

Code Style

  • Use clear, descriptive names
  • Include token estimates for all commands
  • Provide practical examples
  • Keep descriptions concise but comprehensive

Testing

  • Test commands in Claude Code before submitting
  • Verify token estimates are reasonable
  • Ensure workflows coordinate properly
  • Check for conflicts with existing commands

Commit Messages

  • Use clear, descriptive commit messages
  • Format: type: description
  • Types: feat, fix, docs, refactor, test

Example:

feat: add database-architect command
fix: update token estimates for visual-designer
docs: add examples to automation-engineer

Pull Request Process

  1. Fork the repository
  2. Create a feature branch (git checkout -b feature/new-command)
  3. Make your changes
  4. Test thoroughly in Claude Code
  5. Commit with clear messages
  6. Push to your fork
  7. Submit a pull request with:
    • Description of changes
    • Testing performed
    • Token usage verification

Command Naming Conventions

  • Use lowercase with hyphens: visual-designer
  • Be descriptive but concise
  • Avoid abbreviations unless widely known
  • Group related commands with common prefixes

Token Estimation Guidelines

  • Small tasks: 500-1,500 tokens
  • Medium tasks: 1,500-3,000 tokens
  • Large tasks: 3,000-5,000 tokens
  • Complex workflows: 4,000-6,000 tokens

Test actual usage and update estimates accordingly.

Questions?

Open an issue for:

  • Clarification on guidelines
  • Suggestions for improvements
  • Discussion of new features
  • Help with contributions

Thank you for helping make CCGlobalCommands better!