agentleFS
OneWave-AIOneWave-AI/claude-skills/cowork-folder-organizer/SKILL.md

cowork-folder-organizer

Skill378 starsChanged 52 days agoMITView on GitHub

Cowork-style cleanup of a messy folder or shared drive -- classifies every file, proposes a target structure, renames to a consistent convention, deduplicates, and writes an index. Use on Downloads, Desktop, client folders, or team shared drives.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 1 is distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Execute: Create the structure, move and rename files, quarantine duplicates and junk into _review-before-deleting/ -- never delete outright.

    line 22
  2. Very specific:

    Index: Write INDEX.md at the root: the new tree, what lives where, and a table of the 20 most recently active files so the owner can …

    line 23
  3. Very specific:

    Never delete. Quarantine to _review-before-deleting/ and let the human pull the trigger.

    line 28
  4. Very specific:

    Keep names human-readable.

    line 30
  5. Distinctive:

    If the folder is a shared drive, flag files other people may have links to before renaming them.

    line 31

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Cowork Folder Organizer

Turn a chaotic folder into an organized, navigable workspace the way Claude Cowork works through a directory: inventory everything first, propose before moving, then execute with a full undo trail. Never destroy information -- every move is logged and reversible.

Input

Take a directory path. If none is provided, ask for one. Confirm scope before touching anything: subdirectories included or not, and whether any paths are off-limits.

Workflow

  1. Inventory. Walk the directory. For every file record name, extension, size, modified date, and a one-line content summary (read headers/first lines for documents, metadata for media). Do not open files larger than 50MB; classify by name and type only.
  2. Classify. Group files into categories inferred from the actual contents -- not a generic template. Typical clusters: active project files, reference material, financial documents, contracts, media/assets, installers and archives, duplicates, junk (temp files, ~$ locks, empty files).
  3. Detect duplicates. Match by checksum first, then by near-identical names with version suffixes (final, final2, v3, (1)). Identify the canonical copy (newest complete version) and mark the rest.
  4. Propose. Present the target folder tree, the naming convention (default: YYYY-MM-DD-descriptive-name.ext for dated documents, client-project-doctype for client work), the duplicate resolution list, and the junk list. Wait for approval before moving anything.
  5. Execute. Create the structure, move and rename files, quarantine duplicates and junk into _review-before-deleting/ -- never delete outright. Log every operation as old-path -> new-path in _organizer-log.md.
  6. Index. Write INDEX.md at the root: the new tree, what lives where, and a table of the 20 most recently active files so the owner can re-orient instantly.

Rules

  • Propose first, move second. Never reorganize without showing the plan.
  • Never delete. Quarantine to _review-before-deleting/ and let the human pull the trigger.
  • Preserve anything that looks like the only copy of original work (contracts, signed PDFs, raw footage) in place if its location seems intentional.
  • Keep names human-readable. No hash suffixes, no ALL_CAPS, no spaces-to-underscores churn on files that are already well named.
  • If the folder is a shared drive, flag files other people may have links to before renaming them.
  • Report at the end: files processed, moved, renamed, duplicates found, space reclaimable.

Scheduled Mode

When run as a recurring Cowork scheduled task (e.g. weekly on Downloads), skip the approval step only for previously approved rules: apply the established convention to new files, quarantine new junk, and post a summary of what moved. Anything that does not match an existing rule goes to _needs-decision/ and is listed in the summary.

Quick Commands

  • "Organize [path]" -- full workflow with proposal
  • "Just show me the plan" -- phases 1-4 only, no changes
  • "Dedupe only" -- duplicate detection and quarantine, no restructuring
  • "Undo" -- reverse the last run from _organizer-log.md

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.