注释
代码注释与标注
cruft-sweep
parsoFish
Rule-based, directory-by-directory cleanup of generated/abandoned/scratch cruft from the forge tree. Every removal is justified by an EXPLICIT rule, never a hardcoded path, so the rule set stays auditable and adjustable. Reports a plan before deleting; deletes nothing matched by a SAFETY exclusion.
codemod
camshaft
How to structurally search and rewrite Cadenza programs with the cdz query / rewrite codemod tool, and how to ask the COMPILER for a semantic fact with cdz type / cdz uses / cdz check (all in the unified cdz binary). Read this whenever the task is finding or transforming code by SHAPE rather than text — structural search-and-replace, a rename/peephole/wrap refactor, a multi-rule simplifier pass, running a codemod across files/directories (apply in place, diff preview, or JSON), structurally diffing two programs (which subtrees changed), finding duplicated subtrees (exact or near-clone / anti-unification), counting occurrences of a form, extracting spans of matching nodes, or building on the query/Tree matcher API — OR when the task is a SEMANTIC query the shape layer can't answer: the type of a definition (cdz type), every source location that references a name (cdz uses, a span-mapped go-to-references), every well-formedness fault (cdz check, "diagnostics as you type"), a name's definition (cdz def, go-to-definition), the bindings visible at a point (cdz scope, variable scope tracking), a module's exported interface (cdz exports), a definition's/built-in's documentation (cdz doc / cdz doc-at), or a definition's compilation DISPOSITION and concrete monomorphizations (cdz instantiations — whether a def was inlined, specialized, emitted as a standalone function, or transformed into an accumulator loop, and at which concrete types / dictionaries a generic was specialized). Covers the ,x/,@xs pattern language, structural guards (is-literal/head-is/matches/not), relational context (inside/has), multi-rule sets + traversal strategy, multi-file/--write/--diff/--json, the diff (structural tree-diff) and clones (content-hash duplicate + --near anti-unification) subcommands, lint mode (anti-pattern checker / CI gate), the semantic type/uses compiler queries, the COMBINED cdz query --where 'type-of(x) = T' (a structural match filtered by a compiler type predicate), the CLI, the library API, and the self-hosted sidecar map.
pr-review-loop
xpepper
Use when addressing open PR review comments from any reviewer (human or bot) within the current agent session. For a fresh-context-per-comment approach, use ralph-wiggum-loop instead.
orchestrate-issues
picoduck
Coordinate explicitly requested Wollipog child-session issue campaigns through merge, cleanup, recursive follow-ups, and archival. Use only when the user invokes this skill or explicitly asks to orchestrate or delegate issue implementation across sessions. A request to claim, implement, or fix multiple issues alone stays in the current session and does not trigger this skill. Do not trigger merely because the issues concern Orchestrator features.
luau-type-expert
dig1t
Professional Luau type-checking and clean code specialist for Roblox development. Use this skill when: - Writing or reviewing Luau code that needs proper type annotations - Fixing type errors from luau-lsp or luau-analyze - Converting untyped Lua/Luau to strictly typed code - Designing type-safe APIs, modules, or data structures - Understanding Luau type system features (generics, unions, intersections, refinements) - Optimizing code for luau-lsp compatibility - Setting up --!strict mode compliance - Creating type definitions (.d.luau files) - Debugging "Type X is not compatible with Y" errors - Writing metatables with proper type support Triggers: "type error", "type annotation", "luau types", "strict mode", "--!strict", "type checking", "luau-lsp", "type mismatch", "generic type", "union type", "type narrowing", "type refinement", "type cast", "export type", "typeof"
docx
evolving-machines-lab
"Comprehensive document creation, editing, and analysis with support for tracked changes, comments, formatting preservation, and text extraction. When Claude needs to work with professional documents (.docx files) for: (1) Creating new documents, (2) Modifying or editing content, (3) Working with tracked changes, (4) Adding comments, or any other document tasks"
skill-installer
alffei
Install skills into the global or workspace-specific skills directory from a curated list or a GitHub repo path. Use when a user asks to list installable skills, install a curated skill, or install a skill from another repo (including private repos).
code-review-typescript
Jamie-BitFlight
Provides TypeScript-specific code review patterns covering strict mode, ESM, type safety, branded types, discriminated unions, async patterns, runtime safety, and common anti-patterns. Activates on detection of tsconfig.json, .ts, or .tsx files during code review — loaded automatically by dh:code-reviewer.
ugc-safety-features
cruisediary
Detects missing report, block, filter, and contact information mechanisms required for all apps with user-generated content under Guideline 1.2.
gh-address-comments
data219
Help address review/issue comments on the open GitHub PR for the current branch using gh CLI; verify gh auth first and prompt the user to authenticate if not logged in.
codebase-documenter
alunadev
Create comprehensive, beginner-friendly documentation for codebases. Use when writing READMEs, architecture guides, code comments, or API documentation.
linear
data219
Manage issues, projects & team workflows in Linear. Use when the user wants to read, create or updates tickets in Linear.
native-thought-execution
seeker71
"Run a thought on the body's own host resources before renting a frontier mind. This is form-first-reasoning as an invokable door: classify the thought, ask the sovereign body (native router → local lattice → public read door), and ROUTE — a grounded structural hit relays attributed to its NodeID and STOPS at near-zero cost; a genuine miss is the 'yes' that escalates to remote reasoning. The decision is a four-way-proven recipe (form-first-router.fk → 511 on Go/Rust/TS/fkwu), not a vibe. Use this skill whenever a thought could be answered from this repo's grounded body — structural questions (what-is, shape-of, NodeID-of, equivalence, annotate), 'is this thought mine or am I renting it', work offline / air-gapped, or to make the sovereignty gate explicit on a query. Triggers on: native thought, think on the body, form-first, sovereignty gate, ask the body first, is this thought mine, run this thought natively, host-resource-native, answer offline, NodeID it, before spending the rented mind."
rota-brainstorm
l4ci
Use when a Major feature or P0 bug needs design negotiation before implementation planning, or on "brainstorm 42", "design this item".
rota-review
l4ci
Use on "review this", "check before I ship", "look over the branch", before a merge or PR, or implicitly from /rota-ship.
cognitive-task-analysis
curiositech
Methodology for eliciting expert knowledge about cognitive processes, decisions, and mental models in complex tasks
validate-implementation-plan
b-mendoza
Audit and annotate an AI-generated implementation plan for requirements traceability, YAGNI compliance, and assumption risks. Use when reviewing, validating, or auditing an implementation plan or design proposal produced by an AI agent.
rota-plan
l4ci
Use when an item or milestone slice is too big to one-shot, or when alignment matters before code lands, or on "plan 42", "write a plan for this".
atlcli
data219
Use when the user asks to work with Atlassian Jira or Confluence from the terminal using atlcli, including authentication profiles, Jira issue/sprint/worklog operations, Confluence wiki/docs sync, inline comments, troubleshooting, and JSON output for automation.
github-pr-agent-review-loop
data219
Use when a GitHub pull request has Codex or CodeRabbit review-agent findings and the user wants the review loop handled end to end
symfony-doctrine-fetch-modes
data219
Optimize Doctrine queries with fetch modes, lazy loading, extra lazy collections, and query hints for performance
apcball
apcball
Review Checklist - 快速参考清单
symfony-doctrine-relations
data219
Define Doctrine entity relationships (OneToMany, ManyToMany, ManyToOne); configure fetch modes, cascade operations, and orphan removal; prevent N+1 queries
receiving-code-review
data219
Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation