taste-check_skill

This skill reviews code with Linus Torvalds' good taste philosophy, eliminating defensive patterns, deep nesting, and runtime mutations to improve quality.

2.4k

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill kingkongshot/pensieve --skill taste-check

  • SKILL.md4.7 KB

Overview

This skill reviews code using Linus Torvalds' "good taste" philosophy to find defensive patterns, special cases, deep nesting, and other code smells. It provides a compact taste score, highlights fatal issues, lists code smells, and gives concrete refactor suggestions with before/after examples. Use it to harden code quality, simplify logic, and prioritize refactors.

How this skill works

I inspect the provided files, diffs, or pasted snippets and run a checklist based on ten taste principles: trust upstream data, eliminate special cases, avoid post-return mutation, ban ad-hoc string concatenation, enforce single responsibility, limit nesting, extract pure functions, validate early, fail fast, and delete dead code. The output is a structured review with a taste score, top problems, detailed code-smell checklist, prioritized fixes, and concrete code examples for remediation.

When to use it

  • Code review for quality and maintainability
  • Refactoring candidate assessment
  • Checking PR diffs for new code smells
  • Validating architecture and data flow assumptions
  • Teaching teams a stricter taste-driven standard

Best practices

  • Trust upstream data; validate once at the boundary and avoid downstream defensive defaults
  • Prefer early return and small pure functions to keep nesting shallow
  • Return fully-formed results; avoid modifying returned objects after the fact
  • Assemble strings/CSS in one place; avoid incremental concatenation across code paths
  • Keep functions focused (ideally <50 lines); split responsibilities into testable pure functions
  • Fail fast on invalid data instead of hiding issues with fallbacks or try-catch

Example use cases

  • Review a single file or module for multiple code smells and get exact refactor steps
  • Analyze a git diff to decide whether recent changes introduced defensive code or special-case hacks
  • Inspect pasted code snippets to produce quick before/after rewrites and a taste score
  • Prioritize technical debt: produce a short ranked list of refactors for immediate action
  • Audit string-building or CSS generation logic to centralize and eliminate concatenation

FAQ

Specify a file path, paste a code snippet, or provide a git diff. I will analyze the chosen scope and report findings.

Will it auto-apply fixes?

No. It provides concrete before/after examples and prioritized refactor steps for you to apply manually or script into your tooling.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
taste-check skill by kingkongshot/pensieve | VeilStrat