- [Overview](#overview)
/plugin marketplace add athola/claude-night-market/plugin install conserve@claude-night-marketThis skill inherits all available tools. When active, it can use any tool Claude has access to.
modules/code-bloat-patterns.mdmodules/documentation-bloat.mdmodules/git-history-analysis.mdmodules/quick-scan.mdmodules/static-analysis-integration.mdSystematically detect and eliminate codebase bloat through progressive analysis tiers. Reduces context consumption, improves navigability, and identifies technical debt hotspots.
Bloat Categories:
# Heuristic-based detection
/bloat-scan
Verification: Run the command with --help flag to verify availability.
Detects:
Output: Prioritized quick wins with token savings estimates
# Focus on specific bloat types
/bloat-scan --level 2 --focus code
/bloat-scan --level 2 --focus docs
/bloat-scan --level 2 --focus deps
Verification: Run the command with --help flag to verify availability.
Detects:
Output: Detailed findings with confidence levels, remediation steps
# Deep analysis across all dimensions
/bloat-scan --level 3 --report audit-report.md
Verification: Run the command with --help flag to verify availability.
Detects:
Output: Full audit report, refactoring roadmap, automated fixes
Recommended Triggers:
Avoid Using:
| Level | Confidence | Action |
|---|---|---|
| HIGH (90-100%) | Multiple signals confirm | Safe to remove |
| MEDIUM (70-89%) | Single tool + heuristic | Review before action |
| LOW (50-69%) | Heuristic only | Manual investigation needed |
| INVESTIGATE | Edge case | Domain knowledge required |
Priority Score = (
(Token_Savings × 0.4) +
(Maintenance_Impact × 0.3) +
(Confidence_Level × 0.2) +
(Fix_Ease × 0.1)
)
Verification: Run the command with --help flag to verify availability.
Output ranked by descending priority score.
All above + deep cross-file analysis, dependency graphs, automated fixes
=== Bloat Detection Report ===
Scan Level: 1 (Quick Scan)
Duration: 2m 34s
Files Scanned: 847
HIGH PRIORITY (Immediate Action):
❌ plugins/old-plugin/legacy.py
- 1,247 lines (God class)
- Unchanged 18 months
- No imports found (dead code)
- Confidence: 95%
- Token Savings: ~4,500 tokens
- Action: DELETE or archive
❌ docs/archive/old-guide.md
- 87% similar to docs/current-guide.md
- Duplicate content detected
- Confidence: 92%
- Token Savings: ~2,100 tokens
- Action: MERGE or remove
MEDIUM PRIORITY (Review Soon):
⚠️ src/core/manager.py
- 634 lines, 15 methods
- God class candidate
- Confidence: 75%
- Impact: High (core module)
- Action: REFACTOR into smaller classes
⚠️ 5 files with duplicate error handling
- Pattern detected across modules
- Confidence: 70%
- Token Savings: ~1,200 tokens
- Action: EXTRACT to shared utility
STATS:
- Estimated bloat: 3,847 lines (14% of codebase)
- Potential token savings: ~18,000 tokens
- Context reduction: ~12%
- High-confidence findings: 3
- Medium-confidence findings: 6
NEXT STEPS:
1. Review HIGH priority items (3 findings)
2. Run Tier 2 scan for detailed analysis: /bloat-scan --level 2
3. Create cleanup branch: git checkout -b cleanup/bloat-reduction
4. Track progress with issue: /create-issue "Bloat reduction Q1 2025"
Verification: Run git status to confirm working tree state.
NEVER auto-delete without approval:
--dry-run flag for all operationsAutomatic Cache Directory Exclusion:
All scan operations automatically exclude files and directories using a three-tier approach:
Default Excludes (always applied):
# Python
.venv/, venv/, __pycache__/, .pytest_cache/, .mypy_cache/, .ruff_cache/, .tox/
# JavaScript/Node
node_modules/
# Build artifacts
dist/, build/, *.egg-info/
# Version control
.git/
# Dependencies
vendor/
# IDE
.vscode/, .idea/
Verification: Run pytest -v to verify tests pass.
.gitignore Integration (if present):
.gitignore.bloat-ignore (optional, bloat-specific overrides):
.gitignore patternsExclusion Priority: Default → .gitignore → .bloat-ignore
False Positive Mitigation:
.gitignore pattern inheritance.bloat-ignore file supportmodules/quick-scan.md for heuristic patternsmodules/git-history-analysis.md for churn metricsmodules/code-bloat-patterns.md for anti-patternsmodules/documentation-bloat.md for similarity detectionmodules/static-analysis-integration.md for tool setupcontext-optimization: MECW principles, token reductionperformance-monitoring: Resource usage correlationresource-management: Broader optimization strategiesbloat-auditor: Executes scans, generates reports, recommends actionsCommand not found Ensure all dependencies are installed and in PATH
Permission errors Check file permissions and run with appropriate privileges
Unexpected behavior
Enable verbose logging with --verbose flag
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.