Generate per-file metadata manifests and apply deletions to filter documentation.
npx skills add https://github.com/jasonz-ncc42/local-context7 --skill filter-docs
Please help me install this Agent Skill. Skill: filter-docs Source: https://github.com/jasonz-ncc42/local-context7/tree/main/.claude/skills/filter-docs Command: npx skills add https://github.com/jasonz-ncc42/local-context7 --skill filter-docs
AI-assisted documentation filtering to keep only files useful for developers, removing clutter and preserving essential references.
Run the filter to generate a review manifest, review its suggestions, create a _delete.txt of files to remove, and re-run the filter to apply deletions.
High-intent search queries and answers about installing and using this skill.
Filtering documentation to keep useful files involves generating a per-file metadata manifest with path, size, line count, and a short preview to inform decisions. You review this manifest, mark files to delete, and apply changes to remove deselected files from your repository.
To prune legacy docs for AI-assisted coding context, generate a review manifest to evaluate each file's usefulness, create a deletions list of non-useful files, and apply the deletions to clean up the repository and maintain a lean documentation set.
Removing clutter from mixed-content repositories uses a two-step process: first generating an output review manifest with per-file metadata, then applying changes to remove deselected files based on a created deletions list.
Yes, you need jq installed to generate the review manifest and deletions list. The filtering process relies on this dependency to create per-file metadata and apply the workflow to remove non-useful files from your workspace.
Yes, you can review documentation suggestions before applying deletions. The workflow is iterative: you review the generated manifest, mark files to delete in a _delete.txt file, then re-run the filter to apply the deletions and update the workspace.
The best way to maintain lean documentation sets is using an iterative workflow that generates per-file metadata, allows you to review and mark non-useful files for deletion, then applies those deletions to clean up the repository and preserve essential references.
92Bilal26Automate Docusaurus site deployment to GitHub Pages with GitHub Actions.
atmigtncaRemove AI-generated code slop with a regression-safe, deletion-first cleanup workflow.
VanLReview code differences, test for vulnerabilities, and answer coding questions.
majesticlabs-devGenerate categorized Markdown solution files with YAML frontmatter and schema validation.
