Does it permanently delete files immediately?

The skill attempts to move files to the system Trash (on macOS and Linux) rather than using immediate permanent deletion whenever possible, providing an extra layer of safety.

Can I see what will be deleted before it happens?

Yes. The skill provides a comprehensive cleanup plan showing exactly which files are candidates for removal and how much space will be recovered before asking for confirmation.

Will this skill delete my downloaded research papers?

No. All PDF files in the 'papers/' directory are part of the protected list and are never deleted by the cleanup utility.

Is it safe to use this skill on active research projects?

While the skill is designed for completed sessions, its conservative cleanup logic and confirmation prompts make it safe. However, it is best used after the consolidation phase.

What happens if a JSON file is accidentally corrupted?

The skill includes a verification step that uses 'jq' to ensure all preserved JSON files are valid and readable after the cleanup is performed.

Research Session Cleanup

Name: Research Session Cleanup
Author: kthorn

bykthorn

•

データサイエンスとML

Safely removes intermediate files and temporary data from completed research sessions while preserving critical findings and source papers.

概要

Research Session Cleanup is a specialized utility for the Claude Code research workflow that automates the organization and pruning of session directories. It identifies and removes non-essential intermediate search results, logs, and temporary system files, significantly reducing storage footprint before archiving or sharing. Designed with a safety-first philosophy, the skill strictly adheres to a protected list of core assets—such as summaries, JSON databases, and downloaded PDFs—and requires explicit user confirmation before any deletion to ensure no critical research data is lost during the optimization process.

主な機能

Intelligent handling of large cache files with options for compression or archival
3 GitHub stars
Automated identification of safe-to-delete intermediate files and raw search results
Strict preservation of core research outputs including SUMMARY.md and downloaded PDFs
Interactive dry run mode with detailed cleanup analysis and space recovery estimates
Post-cleanup integrity checks to verify JSON validity and essential file presence

ユースケース

Optimizing research folders for archiving or moving to long-term storage
Preparing research data for sharing with collaborators by removing local metadata and logs
Managing disk space by identifying and removing redundant intermediate search artifacts

概要

主な機能

Intelligent handling of large cache files with options for compression or archival
3 GitHub stars
Automated identification of safe-to-delete intermediate files and raw search results
Strict preservation of core research outputs including SUMMARY.md and downloaded PDFs
Interactive dry run mode with detailed cleanup analysis and space recovery estimates
Post-cleanup integrity checks to verify JSON validity and essential file presence

ユースケース

Optimizing research folders for archiving or moving to long-term storage
Preparing research data for sharing with collaborators by removing local metadata and logs
Managing disk space by identifying and removing redundant intermediate search artifacts