重构
代码重组与优化
stage-plan
cxyliuy
Convert vague or high-level requests into a concrete engineering plan with goals, constraints, steps, success criteria, and validation. Stop and wait for human confirmation.
refactor-archeology
ShotaIuchi
Code history and rationale investigation for refactoring. Apply when researching git blame, commit history, code comments, hidden constraints, and historical design decisions.
iterate-skill
maxoreric
迭代改进 Skill 的 Skill。当评价不通过或需要优化时触发。触发词：迭代、改进、优化、iterate、improve。
speed-loop
toomingos
"Systematically optimizes a data processing flow through iterative speed loops. Use when the user wants to improve performance of a data pipeline, query, or processing endpoint. Invoke with the flow to optimize: /speed-loop path/to/file or /speed-loop describe the flow"
safe-refactor-guard
Envy-7z
Enforce ultra-safe refactoring rules to prevent architecture changes, over-engineering, or unintended side effects.
context-ios
ShotaIuchi
iOS development context. Apply when working with .swift files, Xcode project, SwiftUI views, async/await, Combine, MVVM patterns.
mode-awareness
jamelna-apps
When explicitly switching modes or when mode detection shows low confidence. Provides guidance on optimal approach for each detected mode (debugging, performance, feature, refactor, exploration, infrastructure).
refactor
mguinada
"TDD-based code simplification that preserves behavior through tests. Use Red-Green-Refactor cycles to simplify code one test-verified change at a time. DISTINCT FROM: General code review or AI rewriting—this skill requires existing tests and only proceeds when tests confirm behavior is preserved. PROACTIVE: Auto-invoke when test-covered code has complexity (functions >50 lines, high cyclomatic complexity, duplication) and user wants to simplify it safely. Trigger phrases: 'clean up code', 'make code simpler', 'reduce complexity', 'refactoring help'. NOT FOR: Adding features or fixing bugs—use /tdd skill instead."
karpathy-guidelines
8-Sync-Dev
Use this skill before EVERY non-trivial coding task. It enforces Andrej Karpathy-style engineering discipline — read-before-write, test-before-refactor, small steps, boring-is-better, delete-more-than-you-add. Apply whenever the user asks for code, refactor, debug, or review work; the AI MUST cite a rule from this skill before claiming "done".
full-output-enforcement
paulbryantbaker-art
Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.
interactive-task
AgustinAlbonico
Aclarar y ejecutar tareas de desarrollo no-bug (agregar, cambiar, refactorizar, configurar, optimizar) con preguntas adaptativas via question (opciones + respuesta libre). Usar cuando el pedido llega incompleto o ambiguo y se necesita definir alcance rapido.
cleanup
hrdtbs
Investigate and remove unused code, files, and dependencies to keep the codebase clean and maintainable. Use this when you want to "remove unused code", "clean up dependencies", "find dead code", or "optimize project size".
karpathy-guidelines
MILTONADINA
Behavioral guidelines that reduce common LLM coding mistakes. Apply whenever writing, modifying, or reviewing code. Don't assume requirements, don't overcomplicate, don't refactor unrelated code, and define success criteria before implementing. Use this skill on every non-trivial coding task to avoid bloat, silent assumptions, and completion-bias.
clean-codejs-functions
DamianWrooby
Function design patterns emphasizing single responsibility and clarity.
Visual Storyteller
anorbert-cmyk
Creates narrative-driven visuals to explain value and increase conversion, optimizing for clarity and trust.
Code Sculptor
smouj
Automated code refactoring and optimization engine for maintainability and performance
asking-questions
arielperez82
Guidance for asking clarifying questions when user requests are ambiguous, have multiple valid approaches, or require critical decisions. Use when implementation choices exist that could significantly affect outcomes.
fix
wintree86
Backlog manager -- track bugs, improvements, and tech debt in backlog.md. Backlog 관리 스킬 (버그, 개선, 기술 부채). Triggers: /fix, "backlog", "백로그", "bug report", "버그 등록", "개선 등록", "not working", "issue", "problem", "performance", "slow", "upgrade"
product-tower
mrd-bdsmetro
Framework 13 tầng giúp build sản phẩm bài bản. Từ thị trường → user → PMF → features → ship. Tự động route đến sub-skill phù hợp dựa vào giai đoạn hiện tại. Kích hoạt khi user nói "product tower", "build product", "product plan", "product strategy", "build cái gì trước", "nên ship gì", "product roadmap", "validate idea", "tầng mấy", hoặc cần framework toàn diện cho product development.
trellis-baking
simnova
"Bake a 3D asset from a conditioning image through TRELLIS on this machine. Load BEFORE running any bake, choosing sampler parameters, or diagnosing a bad TRELLIS output - carries the measured 15-knob parameter surface, which knobs the repo can and cannot set, the stage-to-defect mapping, and the cost of a variant."
refactoring-agents
arielperez82
Guides ecosystem-level refactors of agents in this repo. Use when agents
clean-transient-comments
iamkaf
Use when the user asks to "clean transient comments", "remove temporary comments", or remove comments that describe past changes rather than documenting current code.
w05-task-closure
qiao-925
任务收尾规范，包含日志生成和优化分析。适用于主要交付完成后的收尾阶段。
refactor
mczkzk
Perform refactoring on <target>