John Masterclass Cc: Coding Agent Conversation Logs from DataClaw

John Masterclass Cc contains fewer than 1,000 conversation logs from AI coding agents, exported by gutenbergpbc in early 2026. The collection captures text-based interactions and tool-use sequences from assistants such as Claude Code and Gemini CLI. It was generated using the DataClaw tool to provide an open alternative to proprietary model interaction data.

Use Cases

• Analyzing 'tool-use' sequences to understand how agents interact with CLI environments
• Training models on 'conversations' between users and coding assistants to improve dialogue flow
• Benchmarking 'text generation' capabilities for specific coding tasks recorded in the logs

Strengths

• Includes 'tool-use' metadata for agentic workflow analysis
• Released under the permissive MIT license
• Captured from modern coding interfaces including Claude Code and Gemini CLI

Limitations

• Small sample size of fewer than 1,000 records
• Lack of explicit column definitions or schema documentation
• Potential for idiosyncratic data due to its origin as a performance art project

Provenance

Source

gutenbergpbc

Collection Method

Exported via DataClaw from AI coding agent sessions

Freshness

Updated as of February 2026.

Geography

United States