How does Coding by Logging differ from standard chat?

Unlike standard chat, it creates a persistent, structured log file where every issue is tracked with severity, specific locations, user feedback, and Claude's specific actions, making it easier to audit and follow over time.

What do the '> JL' and '>> CC' markers mean?

'> JL' represents User instructions or feedback (e.g., 'please fix'), while '>> CC' indicates Claude's concise response and confirmation of the action taken (e.g., 'DONE').

Can I use this for architecture discussions?

Yes, the skill is designed for architecture discussions, large refactors, and any multi-step development process requiring a record of why specific decisions were made.

Does it create massive, unreadable documentation files?

No, the skill enforces a 'concise and readable' rule, prioritizing tables, bullet points, and brief 1-2 sentence responses over verbose explanations to ensure logs remain actionable.

Coding by Logging

Name: Coding by Logging
Author: jluo41

byjluo41

0•

コラボレーションツール

Facilitates structured code reviews and iterative improvements through documented AI-human dialogue in persistent session logs.

Coding by Logging transforms the standard ephemeral AI interaction into a formal, session-based code review process. By utilizing a structured log file to track findings, user instructions, and AI actions, it ensures that every architectural decision and bug fix is documented and traceable. This skill is particularly valuable for complex refactors or multi-step changes where maintaining a clear history of intent and execution is critical for project maintainability.

主な機能

01Interactive human-in-the-loop feedback cycle using specific shorthand markers

02Concise table-based reporting to minimize documentation overhead and maximize readability

030 GitHub stars

04Automated change summaries and status tracking for deferred or completed tasks

05Severity-coded issue logging from Critical to Low levels

06Session-based tracking with timestamps for historical context

ユースケース

01Tracking progress and architectural decisions in complex, multi-day bug fixing sessions

02Conducting asynchronous security or code quality audits with documented resolutions

03Managing large-scale codebase refactors across multiple files and directories

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jluo41/research-skills coding-by-logging

For use in Claude.ai and ChatGPT

Download Skill