What is Agent Trace for Claude Code?

Agent Trace is a skill that implements the Agent Trace specification to track whether code was written by a human or an AI model, providing full transparency into code origin.

Does it support different AI models?

Yes, it captures specific model metadata (e.g., Claude 3.5 Sonnet vs. Claude 4) so you can compare the quality and volume of contributions from different AI versions.

How is the attribution data stored?

Trace records are stored locally in your project's `.agent-trace/` directory as JSON files, ensuring that your attribution data remains private and resides with your code.

Does Agent Trace track human edits to AI code?

Yes, it supports multiple contributor types including 'ai', 'human', and 'mixed' to accurately reflect collaborative editing between developers and AI agents.

Can I use this for legal and compliance auditing?

Yes, the skill includes a dedicated report generation feature specifically designed to provide the documentation needed for compliance audits and IP verification.

Agent Trace

Name: Agent Trace
Author: taiyousan15

bytaiyousan15

•

Analíticas y Monitorización

Tracks and attributes AI-generated code to maintain transparency, compliance, and contribution analytics within your codebase.

Agent Trace is a specialized capability for Claude Code designed to monitor and document the origin of every code change. Based on the Agent Trace specification (v0.1.0), it automatically captures metadata including contributor type (AI vs. human), specific model versions, and timestamps during the development process. This skill is essential for teams requiring clear code attribution for legal compliance, quality auditing, or analyzing AI productivity and contribution ratios across their repositories.

Características Principales

01Detailed trace logs capturing file paths, contributor types, and Git context

02Compliance-ready report generation for code origin and provenance verification

031 GitHub stars

04Real-time AI vs. human contribution statistics and model-specific analytics

05Automated archiving and management of trace data within the local .agent-trace directory

06Searchable history for specific file edits to identify exactly who or what wrote the code

Casos de Uso

01Measuring the impact and productivity of different AI models across a development team

02Identifying the specific AI model or developer responsible for a bug or complex logic implementation

03Auditing codebases for legal and copyright compliance regarding AI-generated content

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add taiyousan15/taisun_agent agent-trace

For use in Claude.ai and ChatGPT

Download Skill