storage-reclaim
Rapidly find and reclaim disk storage by identifying build artifacts, git garbage, temp files, and other space hogs. Use when disk is full or running low on space.
When & Why to Use This Skill
The Storage Reclaim skill is a high-performance utility designed to rapidly identify and eliminate disk-space hogs such as build artifacts, Git garbage, and temporary system files. By automating complex directory investigations and providing safe cleanup patterns for major ecosystems like Rust, Node.js, and Python, it helps developers maintain optimal system performance and recover gigabytes of storage efficiently.
Use Cases
- Development Environment Cleanup: Automatically locating and removing massive 'node_modules' or Rust 'target' directories across multiple projects to reclaim significant disk space.
- Git Repository Maintenance: Identifying and purging 'tmp_pack' files and other git garbage resulting from interrupted operations or large clones.
- Virtual Environment Management: Finding and deleting stale Python virtual environments and hidden cache directories like '__pycache__' that accumulate over time.
- System-Level Optimization: Streamlining the cleanup of Docker images, containers, and Homebrew caches to resolve 'disk full' errors during critical workflows.
| name | storage-reclaim |
|---|---|
| description | Rapidly find and reclaim disk storage by identifying build artifacts, git garbage, temp files, and other space hogs. Use when disk is full or running low on space. |
| version | 1.0.0 |
Storage Reclaim
Rapid parallel investigation and cleanup of disk storage.
Quick Start
# Top-level overview
du -sh /path/*/ 2>/dev/null | sort -hr | head -20
# Drill into specific directory
du -sh /path/subdir/*/ 2>/dev/null | sort -hr | head -15
Common Space Hogs
1. Rust Build Artifacts (target/)
- Location: Any Rust project root
- Size: 1-10+ GB per project
- Safe to delete: Yes (rebuilds on next
cargo build)
# Find all Rust target directories
find ~ -type d -name "target" -exec du -sh {} \; 2>/dev/null | sort -hr | head -20
# Clean specific project
rm -rf /path/to/project/target
# Or use cargo
cd /path/to/project && cargo clean
2. Git Garbage (tmp_pack files)
- Location:
.git/objects/pack/tmp_pack_* - Cause: Interrupted git operations
- Size: Can be gigabytes
# Check for git garbage
git count-objects -vH
# Look for "size-garbage" line
# Remove stale pack files
rm -f .git/objects/pack/tmp_pack_*
# Verify cleanup
git count-objects -vH
3. Node Modules
- Location:
node_modules/in JS projects - Size: 100MB - 2GB per project
# Find all node_modules
find ~ -type d -name "node_modules" -prune -exec du -sh {} \; 2>/dev/null | sort -hr
# Remove (can reinstall with npm install)
rm -rf /path/to/project/node_modules
4. Python Virtual Environments
- Location:
.venv/,venv/,env/ - Size: 100MB - 1GB per environment
find ~ -type d \( -name ".venv" -o -name "venv" -o -name "env" \) -exec du -sh {} \; 2>/dev/null | sort -hr
5. Hidden Temp Directories
- Location:
.tmp/,.cache/,__pycache__/ - Often overlooked by
duon directories
# Check hidden dirs specifically
du -sh /path/.* 2>/dev/null | sort -hr | head -10
6. Julia Artifacts
- Location:
~/.julia/artifacts/,~/.julia/compiled/ - Size: Can grow to many GB
du -sh ~/.julia/*/ 2>/dev/null | sort -hr
7. Docker
docker system df
docker system prune -a # Remove all unused images/containers
8. Homebrew
brew cleanup --dry-run # Preview
brew cleanup # Actually clean
Investigation Pattern
- Start broad:
du -sh /path/*/ | sort -hr | head -20 - Drill into largest: Repeat for subdirectories
- Check hidden:
du -sh /path/.* | sort -hr - Git check:
git count-objects -vHin any repo - Clean safely: Remove build artifacts first (always regeneratable)
Safety Rules
- Always safe to delete:
target/,node_modules/,.tmp/,__pycache__/, build/ - Check first:
.git/(might have garbage, might be real history) - Never delete blindly: Actual source code,
.git/objects/pack/*.pack(real packs) - Regeneratable: Anything that
cargo build,npm install,pip installcreates
Parallel Investigation
Run multiple du commands simultaneously for faster discovery:
# In parallel (use separate terminal or background)
du -sh ~/project1/*/ | sort -hr &
du -sh ~/project2/*/ | sort -hr &
wait
Scientific Skill Interleaving
This skill connects to the K-Dense-AI/claude-scientific-skills ecosystem:
Graph Theory
- networkx [○] via bicomodule
- Universal graph hub
Bibliography References
general: 734 citations in bib.duckdb
Cat# Integration
This skill maps to Cat# = Comod(P) as a bicomodule in the equipment structure:
Trit: 0 (ERGODIC)
Home: Prof
Poly Op: ⊗
Kan Role: Adj
Color: #26D826
GF(3) Naturality
The skill participates in triads satisfying:
(-1) + (0) + (+1) ≡ 0 (mod 3)
This ensures compositional coherence in the Cat# equipment structure.