agent-trace

Tracks and reports AI-generated code attribution using Agent Trace specification records.

Updated Mar 23, 2026
One-click install
npx skills add https://github.com/sakamotomomotaro0809-netizen/tateyomi --skill agent-trace-sakamotomomotaro0809-netizen
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-trace
Source: https://github.com/sakamotomomotaro0809-netizen/tateyomi/tree/main/taisun_agent/.claude/skills/agent-trace
Command: npx skills add https://github.com/sakamotomomotaro0809-netizen/tateyomi --skill agent-trace-sakamotomomotaro0809-netizen

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams using AI coding assistants lose visibility into which code was written by humans versus AI models, making compliance audits and contribution analysis difficult. This Skill reads automatically captured trace records and turns them into statistics, searchable history, and compliance reports. ## Core Features & Use Cases - Statistics & Contribution Analysis: View AI versus human contribution rates and per-model breakdowns from stored trace records. - Trace Search & History: List recent traces or find the full edit history for a specific file such as src/api/auth.ts. - Compliance Reporting & Archiving: Generate compliance reports and archive traces older than a configurable number of days (default 30). - Use Case: Before a release audit, run the report command to produce a compliance document showing which files contain AI-generated code, which model produced it, and the associated Git commits. ## Quick Start Ask the agent to show the AI contribution statistics for this project using the agent-trace skill.

Frequently Asked Questions about agent-trace

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check AI code contribution statistics in my project?▼

Run the /agent-trace stats command or call TraceStore.getStatistics() to view AI versus human contribution rates and per-model breakdowns. Statistics are computed from JSON trace records stored in the .agent-trace/traces directory.

How to find which files were written by AI in a codebase?▼

Use /agent-trace find with a file path, or call TraceStore.findByFile() with a path like src/api/auth.ts. Each trace record lists files, contributor type (ai, human, mixed, unknown), the model used, and line ranges.

Does agent-trace record traces automatically?▼

Yes, a PostToolUse hook (agent-trace-capture.js) automatically creates a trace record whenever the AI edits or writes files. Records include file paths, model information, timestamps, and Git commit and branch context.

What files are excluded from AI code tracing?▼

Sensitive and irrelevant paths are excluded by default: node_modules, .git, .agent-trace, .env files, secrets and credentials directories, and *.pem or *.key files. This protects confidential material from being recorded.

How do I archive old agent-trace records?▼

Run /agent-trace archive with an optional day count (default 30). Traces older than the threshold move from .agent-trace/traces into .agent-trace/archive, keeping the active store small while preserving history.