Does Bloat Detector automatically delete my code?

No, Bloat Detector never auto-deletes. It identifies issues and suggests remediations that require your explicit approval. It also supports dry-runs and creates backup branches for safety.

Can it identify bloat created by AI coding assistants?

Yes, it includes specific modules for AI-generated bloat, such as 'vibe coding' signatures, tab-completion redundancy, and hallucinated dependencies.

When is the best time to run a bloat scan?

It is most effective during quarterly maintenance, before major refactoring, or when you notice high context usage (over 30%) interfering with AI performance.

Are there codebases where I shouldn't use this skill?

Avoid using it during active feature development or for time-sensitive bug fixes. It is also less effective for very small projects under 1000 lines of code.

Bloat Detector

Name: Bloat Detector
Author: athola

byathola

•

123

•

Developer Tools

Identifies and eliminates codebase bloat through progressive analysis to reduce technical debt and optimize AI context usage.

Bloat Detector is a systematic analysis tool for Claude Code that helps developers reclaim context and improve maintainability by rooting out unnecessary code and documentation. It employs a tiered analysis strategy—ranging from instant heuristic scans to deep audits—to identify dead code, AI-generated redundancy, and 'lava flow' anti-patterns. Ideal for managing large repositories, this skill provides actionable remediation types and prioritizes tasks based on potential token savings and confidence levels, ensuring your codebase remains lean and efficient.

Key Features

01Tiered analysis levels from quick heuristic scans to deep static audits

02Git history analysis to identify stale files and low-churn hotspots

03Safety-first remediation with dry-run support and automated backup branches

04123 GitHub stars

05Detection of AI-generated 'tab-completion' bloat and documentation slop

06Smart prioritization scoring based on token savings and confidence

Use Cases

01Preparing a codebase for major architectural refactoring or a production release

02Performing quarterly maintenance to clear technical debt and dead code

03Reducing token consumption when repository context usage exceeds 30%

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add athola/claude-night-market bloat-detector

For use in Claude.ai and ChatGPT

Download Skill