Overview
What It Does
Trace To Training Data packages a focused browser & research workflow for an AI agent. Convert evaluation traces and production logs into SFT examples and preference pairs. Use when graded traces or failure examples exist and need to become training data, when applying rejection sampling to model outputs, or when building DPO pairs from passing and failing runs. It is best suited to users who can review the resulting actions and provide only the accounts, files, or command access needed for the task. It is not a substitute for human approval on destructive, financial, security-sensitive, or public-facing actions.
Task ideas
Popular Use Cases
- Research a topic across public sources
- Collect structured information from web pages
- Navigate a browser-driven workflow
- Summarize findings with source links
Installation
Install this Agent Skill
Claude Code
npx skills add https://github.com/wshobson/agents --skill trace-to-training-dataCommands derived from the public GitHub SKILL.md record. Checked 2026-09-02. Review the source before running them.
Before you start
Requirements
| Claude Code | Required / review |
| Codex | Required / review |
| Public SKILL.md source | Required / review |
| Review instructions and requested permissions before installation | Required / review |
| Paid service | Check source |
| Supported system | Check source |
Popularity context
Why It’s Popular
Trace To Training Data is a verified Agent Skill from wshobson with a public SKILL.md, compatible with Claude Code, Codex.
Alternatives
