Night Market Skill — ported from claude-night-market/conserve. For the full experience with agents, hooks, and commands, install the Claude Code plugin.
Systematically detect and eliminate codebase bloat through progressive analysis tiers.
CategoryExamplesCodeDead code, God classes, Lava flow, duplicationAI-GeneratedTab-completion bloat, vibe coding, hallucinated depsDocumentationRedundancy, verbosity, stale content, slopDependenciesUnused imports, dependency bloat, phantom packagesGit HistoryStale files, low-churn code, massive single commits
bash
/bloat-scan
Detects: Large files, stale code, old TODOs, commented blocks, basic duplication
bash
/bloat-scan --level 2 --focus code # or docs, deps
Adds: Static analysis (Vulture/Knip), git churn hotspots, doc similarity
bash
/bloat-scan --level 3 --report audit.md
Adds: Cross-file redundancy, dependency graphs, readability metrics
DoDon'tContext usage > 30%Active feature developmentQuarterly maintenanceTime-sensitive bugsPre-release cleanupCodebase < 1000 linesBefore major refactoringTools unavailable (Tier 2/3)
LevelConfidenceActionHIGH90-100%Safe to removeMEDIUM70-89%Review firstLOW50-69%Investigate
text
Priority = (Token_Savings × 0.4) + (Maintenance × 0.3) + (Confidence × 0.2) + (Ease × 0.1)
Tier 1 (always available):
modules/quick-scan.md - Heuristics, no toolsmodules/git-history-analysis.md - Staleness, churn, vibe coding signaturesmodules/growth-analysis.md - Growth velocity, forecasts, threshold alertsTier 2 (optional tools):
modules/code-bloat-patterns.md - Anti-patterns (God class, Lava flow)modules/ai-generated-bloat.md - AI-specific patterns (Tab bloat, hallucinations)modules/documentation-bloat.md - Redundancy, readability, slop detectionmodules/static-analysis-integration.md - Vulture, KnipShared:
modules/remediation-types.md - DELETE, REFACTOR, CONSOLIDATE, ARCHIVEPatterns that span plugin boundaries or manifest configuration, discovered through ecosystem-wide audits.
alwaysApply Accumulation#Flag plugins with 3+ skills where alwaysApply: true.
Each always-on skill injects its full text into every session,
creating a baseline token floor before the user types anything.
Sum the estimated_tokens fields to report total per-session cost.
Compare hooks declared in plugin.json or openpackage.yml
against entries in hooks.json. A hook present in hooks.json
but absent from the manifest is invisible to the plugin loader
and cannot be audited, versioned, or disabled through normal
plugin management.
Scan skill files for identical multi-line text blocks repeated across 10+ files (e.g., generic troubleshooting sections like "Command not found / Permission errors / Unexpected behavior"). These are copy-paste artifacts that inflate token cost without adding skill-specific value.
Skills loaded into model context gain nothing from HTML-style
Tables of Contents. Detect ## Table of Contents followed by
bulleted anchor-link lists. These waste tokens since
the model reads sequentially, not via hyperlinks.
Compare files on disk in skills/*/modules/ against the
modules: list in each skill's SKILL.md frontmatter. Files
that exist on disk but are not listed in the manifest are
invisible to progressive loading and may be dead weight or
missing from the load path.
Always excludes: .venv, __pycache__, .git, node_modules, dist, build, vendor
Also respects: .gitignore, .bloat-ignore
--dry-run for previewsbloat-auditor agent - Executes scansunbloat-remediator agent - Safe remediationcontext-optimization skill - MECW principles