How this skill is triggered — by the user, by Claude, or both
Slash command
/story-to-ship:code-qualityThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
```
references/INDEX.mdreferences/code-smells.mdreferences/design-principles.mdreferences/oop/cpp/cpp-toolchain.mdreferences/oop/cpp/formatting-rules.mdreferences/oop/cpp/index.mdreferences/oop/cpp/invocation.mdreferences/oop/cpp/naming-tables.mdreferences/oop/cpp/oop-smells.mdreferences/oop/cpp/review-checklist.mdYOU MUST RUN ALL UNIVERSAL CHECKS AND PARADIGM CHECKS BEFORE COMMITTING. No exceptions.
Violating the letter of this rule is violating the spirit of this rule.
Announce at start: "I am using the code-quality skill to [format/lint/review] [description]."
Before every commit:
references/code-smells.md Universal Smells section -- none present in changed codereferences/design-principles.md -- no principle violations (Don't Repeat Yourself (DRY), You Aren't Gonna Need It (YAGNI), Composed Method, Beck's Rules)[+] All 3 pass -> continue to Step 2 [-] Any failing -> fix before proceeding
Run in order; stop at first match:
.cpp or .hpp files modified -> C++ Object-Oriented Programming (OOP) tier (references/oop/cpp/index.md)Python OOP tier (Python files with class definitions): deferred -- references/oop/python/ does not exist yet. Apply universal tier only.
Scripting tier (.sh, .ps1, .bash): deferred -- references/scripting/ does not exist yet. Apply universal tier only.
Load references/[paradigm]/index.md and apply all checks listed there.
[+] All paradigm checks pass -> proceed to commit [-] Any failing -> fix before committing
Readability is more important than development speed or execution speed.
Comments are not intent. When a comment exists to explain what a value means, rewrite the code -- not the comment.
| Signal | Wrong fix | Right fix |
|---|---|---|
// 0=Fullscreen, 1=AutoCOM... above a switch | Add a better comment | Replace magic numbers with a named enum |
magic_value = 42; // timeout in ms | Document the constant | constexpr int kTimeoutMs = 42; |
Function with a bool parameter | Comment at the call site | Replace with an enum or named overloads |
Rule: If a comment explains what a value is, the comment is a code smell. Make the code say it.
| Excuse | Reality |
|---|---|
| "I'll check quality at the end before the PR" | Run universal checks after every meaningful change. Catch issues early. |
| "The paradigm tier does not apply to this small change" | Paradigm tier fires on file extension. No exceptions for small changes. |
| "This code is too simple to need smell checks" | Universal smells (Long Method, Duplicated Code) appear in simple code. Run the check. |
| "The auto-formatter will handle it" | Run your project's formatter explicitly -- it does not run itself. |
| "I'll clean up style in a follow-up PR" | Style debt compounds. Clean now while context is fresh. |
references/code-smells.md -- universal smells (Long Method, Long Parameter List, Duplicated Code, Speculative Generality, Divergent Change, Shotgun Surgery)references/oop/cpp/oop-smells.md -- OOP-specific smells (Feature Envy, Data Clumps, Primitive Obsession, etc.)references/design-principles.md -- design heuristics (all paradigms)references/oop/cpp/index.md -- OOP tier dispatch tablereferences/oop/cpp/cpp-toolchain.md -- clang-format, clang-tidy, cmakereferences/oop/cpp/formatting-rules.md -- human-reviewable C++ formatting patternsreferences/oop/cpp/naming-tables.md -- OOP naming conventionsreferences/oop/cpp/review-checklist.md -- full OOP pre-commit checklistreferences/oop/cpp/invocation.md -- OOP tier invocation instructionsversioning skilltesting skillcpp-patterns skillProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub jpegthedev/story-to-ship --plugin story-to-ship