simplify-codebase

Audit and remove accidental complexity from codebases with evidence-backed deletion.

83.7k|5.7k|Updated Apr 15, 2026
Ready to run
npx skills add https://github.com/tt-a1i/archify --skill simplify-codebase-tt-a1i
💬 Prompt for ChatGPT / Claude Chat▼
Please help me install this Agent Skill.
Skill: simplify-codebase
Source: https://github.com/tt-a1i/archify/tree/main/.agents/skills/simplify-codebase
Command: npx skills add https://github.com/tt-a1i/archify --skill simplify-codebase-tt-a1i

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Codebases accumulate dead code, duplicate state, redundant layers, and ownerless abstractions that increase maintenance burden. This Skill provides a disciplined workflow to prove which code is genuinely removable and execute safe, validated simplifications. ## Core Features & Use Cases - Simplification Audit (Survey mode): Read-only investigation that ranks deletion candidates by confidence, blast radius, and maintenance concepts retired. - Authorized Simplification (Change mode): Proves each cut with consumer maps and contract analysis, removes the obligation vertically, and validates the surviving system in widening verification rings. - Boundary and Lifecycle Analysis: Maps validation, concurrency, and cleanup mechanisms to owners and failure windows before classifying them as redundant. - Use Case: Ask the agent to audit a repository for dead code and over-engineering; it returns ranked proof records showing reachability evidence, consequences, and the smallest check that would expose an incorrect cut. ## Quick Start Use the simplify-codebase skill to audit this repository and rank the safest high-impact simplifications without editing files.

Frequently Asked Questions about simplify-codebase

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find dead code in a large repository?▼

Run a simplification audit in Survey mode, which partitions the system into domains, builds a coverage map, and ranks candidates by reachability evidence. It classifies every consumer as runtime, support, or uncertain rather than relying on search counts alone.

How to safely remove unused code without breaking production?▼

Use Change mode, which traces runtime paths from entrypoints, proves each cut with contract and behavior evidence, then validates in widening rings from residue searches to repository gates. Every batch includes an operation receipt with an undo path.

Does this work with any programming language?▼

Yes, the simplification workflow is language-agnostic and uses repository-native search, compiler and linter output, dependency metadata, and version-control history as discovery instruments. It adapts its coverage domains to each repository's architecture.

When should I not use codebase simplification?▼

Avoid it for general code review, onboarding, style-only refactoring, or performance tuning, which are explicitly out of scope. It also will not delete reachable capabilities or compatibility paths without explicit product authorization.

Why does the audit keep some seemingly unused code?▼

Candidates are kept when dynamic reachability is unresolved, a current design decision still owns the code, or the change merely relocates complexity. Unresolved leads are reported with the exact fact needed to decide them.