code-quality

Assess code changes for reuse, minimality, and correct abstraction altitude.

Updated Jun 28, 2026
One-click install
npx skills add https://github.com/sicambria/talkteach-asr --skill code-quality-sicambria
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: code-quality
Source: https://github.com/sicambria/talkteach-asr/tree/main/.claude/skills/code-quality
Command: npx skills add https://github.com/sicambria/talkteach-asr --skill code-quality-sicambria

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code changes often grow larger than the problem requires: duplicated helpers, unrequested abstractions, speculative config, and logic pitched at the wrong level of abstraction. This Skill provides a structured quality-only review pass that judges a change on fit rather than correctness, catching over-engineering before it calcifies into the codebase. ## Core Features & Use Cases - Reuse Check: Applies a Minimality Ladder (does it need to exist, stdlib, existing helper, dependency, minimal custom code) to detect duplicated or unnecessary new code. - Altitude & Subtraction Analysis: Identifies logic pitched too low (inline duplication) or too high (frameworks for one caller), then asks what breaks if each added construct is removed. - Ranked Findings Report: Produces file:line findings with concrete smaller or reused alternatives, distinguishing must-fix violations from optional polish. - Use Case: Before committing a non-trivial diff that works but feels large or clever, run this pass to confirm the change is the smallest correct solution at the right abstraction level. ## Quick Start Review my current uncommitted changes for over-engineering, duplication, and unnecessary abstractions, and report ranked simplification findings.

Frequently Asked Questions about code-quality

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

FAQPage Schema
How do I check if my code change is over-engineered?▼

Run a structured quality pass that restates the approved problem, checks whether each new construct already exists in the codebase, and asks what breaks if each added parameter, flag, or abstraction is removed. Findings are ranked as must-fix or optional polish.

What is the difference between a code quality review and a code review for bugs?▼

A quality review judges fit: reuse, minimality, and abstraction altitude, without hunting for correctness bugs. Bug hunting belongs to red-team or code-review passes; conflating the two dilutes both, so this Skill deliberately excludes correctness analysis.

When should I run a code quality assessment?▼

Run it before committing a non-trivial change, when a working diff feels large or clever, or when assessing a contributor's branch for reuse before integration. Skip it for changes already known to be minimal fast-path edits.

What is the Minimality Ladder in code review?▼

The Minimality Ladder ranks solutions in order: does the code need to exist at all, can stdlib or platform features handle it, does an existing helper fit, can an installed dependency do it, and only then minimal custom code. Choosing a lower rung for taste is a finding.

Can a code quality review return zero findings?▼

Yes. Unlike a bug review, a clean bill is a valid result for a quality pass. If the change is already minimal, reuses existing patterns, and sits at the right abstraction level, the assessment says so plainly.