danicolms

AI-Human Attribution Tracking Skill

1
0
# Install this skill:
npx skills add danicolms/ai-human-attribution-skill

Or install specific skill: npx add-skill https://github.com/danicolms/ai-human-attribution-skill

# Description

Claude Code skill to manage ai-human attribution percentages, based on https://github.com/mmasias/mmasias/blob/main/ai/etiquetadoEtico.md

# SKILL.md

AI-Human Attribution Tracking Skill

Ethical AI collaboration attribution tracking system for software projects.

Overview

This skill provides a framework for transparently tracking and documenting contributions between human developers and AI assistants (like Claude Code). It ensures ethical attribution while maintaining detailed logs of who did what across the development lifecycle.

Why This Matters

  • Transparency: Clear record of AI vs. human contributions
  • Ethics: Honest attribution of creative and technical work
  • Portfolio Building: Accurate representation of your capabilities
  • Team Collaboration: Shared understanding of project history
  • Client Trust: Clients can see what they're paying for

Quick Start

When this skill is active, Claude Code will:

  1. Ask about attribution before starting significant work
  2. Track contributions using the Principal/Support/Equal/None framework
  3. Document your choices in AI-HUMAN-ATTRIBUTION.md
  4. Update overall project stats after completing work
  5. Maintain detailed logs for every feature/task

Usage

Starting a New Task

When you ask Claude Code to build something, it will:

πŸ“Š Before I begin, let's clarify attribution:

This task will be tracked in AI-HUMAN-ATTRIBUTION.md.
Please describe your role:
- Do you have a specific vision/implementation approach?
- Or should I lead on design/implementation?

[Responses guide the Principal/Support/Equal breakdown]

During Development

Claude Code will:
- Note when it suggests something vs. implements your idea
- Ask for clarification on ambiguous contributions
- Track time estimates for both human and AI effort

After Completion

Claude Code will:
- Update the attribution markdown file
- Calculate overall project percentages
- Maintain detailed log entries

File Structure

your-project/
β”œβ”€β”€ AI-HUMAN-ATTRIBUTION.md    # Main attribution document
β”‚   β”œβ”€β”€ Overall Project Attribution (percentages)
β”‚   β”œβ”€β”€ Role Definitions
β”‚   └── Detailed Contribution Log
└── [your project files]

The Role Framework

Contributions are tracked across 4 levels per development phase:

Level Meaning
Principal Primary driver, decision-maker, creator
Support Helper, executor of others' ideas
Equal Balanced collaboration, joint decisions
None No participation in this phase

Development Phases

Each task is analyzed across 4 phases:

  1. Conceptualization: Ideas, vision, requirements
  2. Analysis: Problem-solving, architecture, research
  3. Implementation: Writing code, configuration, assets
  4. Validation: Testing, review, refinement

Example Entry

### 2025-01-15: User Authentication System

**Phase**: Implementation + Validation

| Role | Human | AI |
|------|-------|-----|
| Conceptualization | Principal - JWT vision, user roles | Support |
| Analysis | Equal - Security considerations | Equal |
| Implementation | Support | Principal - Auth0 integration code |
| Validation | Principal - Testing, edge cases | Support |

**Deliverables**:
- βœ… Login/signup forms
- βœ… JWT token management
- βœ… Protected routes
- βœ… Role-based access control

**Tools Used**: Claude Code (Sonnet 4.5), Auth0 docs
**Human Time**: ~2 hours (design, testing)
**AI Time**: ~45 minutes (implementation, debugging)
**Commit**: `abc1234`

**Notes**:
- Human designed the overall auth flow
- AI implemented Auth0 SDK integration
- Human discovered and fixed token refresh bug

Best Practices

  1. Be Honest: Don't downplay either party's contributions
  2. Be Specific: Document what each party actually did
  3. Track Time: Helps future project estimation
  4. Update Regularly: Don't let attribution get stale
  5. Include Context: Notes explain the "why" behind decisions

Integration with Your Workflow

This skill works alongside:
- Git commits (reference commit hashes)
- Project documentation (README references attribution)
- Code reviews (attribution informs review focus)
- Client reporting (transparent effort breakdown)

Customization

You can adapt this system for:
- Solo developers with AI assistance
- Teams using AI pair programming
- Agencies tracking AI-assisted work
- Open source projects with AI contributions

Files Created

This skill includes:
- SKILL.md (this file) - Skill overview and usage
- TEMPLATE_USER_GUIDE.md - Detailed human-facing guide
- TEMPLATE_API_REFERENCE.md - Technical reference for implementation
- templates/ATTRIBUTION.md.template - Attribution document template


Skill Version: 1.0.0

Last Updated: 2025-01-15
Inspired By: @loscolmebrothers/home project attribution system

License: MIT - Feel free to adapt and share

# README.md

AI-Human Attribution Tracking Skill

Ethical AI collaboration attribution tracking system for software projects

πŸ“¦ What's Inside

This skill provides a complete framework for tracking contributions between human developers and AI assistants (like Claude Code, ChatGPT, Copilot, etc.).

Files

ai-human-attribution-skill/
β”œβ”€β”€ README.md                        # This file
β”œβ”€β”€ SKILL.md                         # Skill overview for Claude Code
β”œβ”€β”€ TEMPLATE_USER_GUIDE.md           # Comprehensive user guide
β”œβ”€β”€ TEMPLATE_API_REFERENCE.md        # Technical specification
└── templates/
    └── ATTRIBUTION.md.template      # Attribution file template

πŸš€ Quick Start

1. Install the Skill

Option A: Claude Code CLI

# Copy skill folder to Claude Code skills directory
cp -r ai-human-attribution-skill ~/.claude/skills/

Option B: Manual Import
- Open Claude Code
- Settings β†’ Skills β†’ Import Skill
- Select SKILL.md from this folder

2. Initialize Attribution in Your Project

# Copy the template
cp ai-human-attribution-skill/templates/ATTRIBUTION.md.template your-project/AI-HUMAN-ATTRIBUTION.md

# Edit the header section with your project details
nano your-project/AI-HUMAN-ATTRIBUTION.md

3. Add to Your README

## Attribution

This project uses AI-assisted development with ethical attribution tracking.
See [AI-HUMAN-ATTRIBUTION.md](./AI-HUMAN-ATTRIBUTION.md) for details.

4. Start Tracking

When working with Claude Code, the skill will:
- Ask about attribution before starting work
- Document contributions as you go
- Update your attribution file automatically

πŸ“š Documentation

For Users

For Developers

Templates

🎯 Key Features

βœ… What It Tracks

  • Who did what across 4 development phases
  • Conceptualization (ideas)
  • Analysis (architecture)
  • Implementation (coding)
  • Validation (testing)

  • Contribution levels

  • Principal (led the work)
  • Support (assisted)
  • Equal (collaborated)
  • None (didn't participate)

  • Metrics

  • Overall project percentages
  • Time estimates (human vs. AI)
  • Phase-by-phase breakdown
  • Deliverable tracking

🎨 The Framework

Based on a proven system used in real projects:

β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚                  CONTRIBUTION BREAKDOWN                     β”‚
β”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€
β”‚  Human: β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘ 65%    β”‚
β”‚  AI:    β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–‘β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‘β–‘ 35%    β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜

🌟 Why Use This?

For Solo Developers

  • Build an honest portfolio
  • Track your AI learning journey
  • Understand AI's impact on your productivity

For Teams

  • Transparent AI usage
  • Fair credit distribution
  • Onboarding documentation

For Agencies/Freelancers

  • Client transparency
  • Justify billing (human vs. AI time)
  • Build trust through honesty

For Open Source

  • Normalize ethical AI usage
  • Share realistic AI workflows
  • Contribute to AI literacy

πŸ“– Example Entry

### 2025-01-15: User Authentication

**Phase**: Implementation + Validation

| Role | Human | AI |
|------|-------|-----|
| Conceptualization  | Principal - JWT vision | Support |
| Analysis           | Equal | Equal |
| Implementation     | Support | Principal - Auth0 code |
| Validation         | Principal - Testing | Support |

**Deliverables**:
- βœ… Login/signup forms
- βœ… JWT token management

**Tools Used**: Claude Code (Sonnet 4.5)
**Human Time**: ~2 hours (design, testing)
**AI Time**: ~45 minutes (implementation)
**Commit**: `abc1234`

**Notes**:
- Human designed auth flow
- AI implemented Auth0 SDK
- Human found and fixed token refresh bug

πŸ”§ How It Works

1. Start a Task

You: "Build a user auth system"

Claude: "πŸ“Š Before I start, let's track attribution..."

2. Collaborate

Claude: "I'll implement Auth0 integration [AI: Principal]"
You: "I'll test the token refresh [Human: Principal]"

3. Document

Claude: Updates AI-HUMAN-ATTRIBUTION.md with:
- Contribution matrix
- Time estimates
- Deliverables
- Commit hash
- Context notes

4. Aggregate

Overall: Human 60%, AI 40%
Phase breakdowns updated

🀝 Contributing

This skill is open for improvement:

  1. Fork the repository
  2. Make your improvements
  3. Submit a pull request

Areas for contribution:
- Additional export formats (JSON, CSV)
- Web UI for visual editing
- CI/CD integrations
- Team support (>1 human)
- Localization

πŸ“„ License

MIT License - Feel free to adapt and share

πŸ™ Credits

Created by: Based on the attribution system from @loscolmebrothers/home

Inspired by: The growing need for ethical AI attribution in software development

Version: 1.0.0
Last Updated: 2025-01-15

πŸ”— Resources


Ready to track your AI collaborations ethically?

Start with SKILL.md πŸš€

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.