Where does the skill save the captured files?

Reviews are saved in the 'reviews/' directory, issues in 'issues/', action items are appended to 'TODO.md', and context is saved to 'Prompt.md'.

Does it support Git integration?

Yes, if a commit hash is detected or provided, the skill will associate the capture with that specific commit, which is especially useful for generating code review documentation.

How does the auto-classification work?

It uses a classification algorithm that scans for specific patterns: commit hashes trigger 'reviews', feature keywords trigger 'issues', and urgent/checkbox patterns trigger 'TODO' items.

Can I force a specific category for the captured content?

Yes, you can override the automatic classification by using the --type flag with arguments like 'issue', 'review', 'todo', or 'prompt'.

What is the capture skill for Claude Code?

The capture skill is a specialized tool that allows Claude to save conversation content directly into project files like TODO.md, Prompt.md, or structured directories for issues and code reviews.

Intelligent Content Capture

Name: Intelligent Content Capture
Author: evmts

byevmts

•

Productividad y Flujo de Trabajo

Automatically classifies and saves conversation content into structured code reviews, issue reports, TODO lists, or project prompts.

Acerca de

The Capture skill for Claude Code streamlines the development workflow by intelligently routing conversation snippets to their appropriate documentation targets. By analyzing keywords, commit hashes, and intent, it automatically formats and saves content into specific directories like reviews/ and issues/, or appends to files like TODO.md and Prompt.md. This ensures that valuable insights, bug reports, and action items generated during an AI coding session are never lost and remain organized without manual filing.

Características Principales

42 GitHub stars
Automatic content classification using a priority-based decision flow
Context preservation via automated Prompt.md updates
Structured markdown template generation for reviews and issues
Flexible CLI support for manual type overrides and metadata assignment
Git-aware commit association for contextual code reviews

Casos de Uso

Turning chat-based feedback into formal code review documents with commit references
Maintaining an organized TODO.md with prioritized action items captured during development
Quickly logging feature requests and project issues without leaving the terminal

Acerca de

Características Principales

42 GitHub stars
Automatic content classification using a priority-based decision flow
Context preservation via automated Prompt.md updates
Structured markdown template generation for reviews and issues
Flexible CLI support for manual type overrides and metadata assignment
Git-aware commit association for contextual code reviews

Casos de Uso

Turning chat-based feedback into formal code review documents with commit references
Maintaining an organized TODO.md with prioritized action items captured during development
Quickly logging feature requests and project issues without leaving the terminal