Use when a substantial request needs verified, non-repetitive output. Triggers on: "์ด๊ฑฐ ์ ๋๋ก ํด์ค", "๊ฒ์ฆ๊น์ง ํด์", "ํ๋ค์ค ๋๋ ค์ค", "run the harness", "do this properly with verification", "6๋จ๊ณ๋ก ์ฒ๋ฆฌํด์ค", "๋ฉํ์คํฌ๋ฆฝํธ๋ก". Not for trivial edits or Q&A.
newkayak12/claude-skills
SkillsMP has collected 125 skills from newkayak12/claude-skills. Open a skill to review its source and details.
- Latest recorded source activity
- SkillsMP catalog refreshed
- skills collected
- 125
- GitHub stars
- 3
- GitHub forks
- 1
Skills in this repository
Showing 40 of 125 collected skills.
Use when turning code, docs, notes, or mixed sources into a small-to-moderate, query-efficient Markdown knowledge base with predictable folders, an AI lookup catalog, source-grounded summaries, and RAG-ready metadata.
Use when building knowledge graphs by extracting source-grounded entities, relationships, properties, evidence, schema, and graph-ready data from code, documents, notes, tickets, or mixed corpora.
Use when answering questions over an existing linked Markdown vault, local SQLite index, ontology, knowledge graph, RAG corpus, or mixed knowledge assets while preserving citations, uncertainty, and retrieval traceability.
Use when turning a body of source material into a queryable knowledge system through iterative graph-like exploration: intake, source frontier, linked Markdown vault, ontology, graph records, RAG corpus, and query surfaces.
Use when designing or refining an ontology for a knowledge system: domain concepts, classes, relationship semantics, constraints, controlled vocabularies, reasoning rules, and mappings to graph/RAG/vault artifacts.
Use when preparing a retrieval-augmented generation corpus from code, documents, notes, tickets, or mixed sources by chunking, adding metadata, preserving citations, and designing retrieval/evaluation checks.
Use when a harness stage must delegate Implement or Test work to local Codex CLI from plugin, repo-local, or embedded installs.
Use when installing the harness into a project. Triggers on: "ํ๋ค์ค ์ค์นํด์ค", "์ด ํ๋ก์ ํธ์ ํ๋ค์ค ์ ์ฉํด์ค", "harness install", "set up the harness here", "๊ฒ์ดํธ ์ผ์ค", "convention ์ค์บํด๋ฉํด์ค". Not for running the engine (use the harness skill).
Use when executing an implementation plan with independent tasks using fresh subagents per task. Triggers on: "๊ณํ ์คํํด์ค", "์๋ธ์์ด์ ํธ๋ก ๊ตฌํํด์ค", "subagent-driven", "plan ์คํ", ํ์คํฌ๋ณ๋ก ์์ด์ ํธ ๋ฐฐํฌํด์ค", "๋ ๋จ๊ณ ๋ฆฌ๋ทฐ๋ก ๊ตฌํ", "ํ์ฌ ์ธ์ ์์ ๊ณํ ์คํ".
Use when a task is large enough that diving in without a plan leads to rework. Triggers: "๊ตฌํ ๊ณํ ์จ์ค", "implementation plan", "์ฝ๋ฉ ์ ์ ๊ณํ ์ก์์ค", "migration plan", "plan before coding", "์คํ ๋ณ๋ก ์ ๋ฆฌํด์ค", "๋ฆฌํฉํ ๋ง ๊ณํ".
Use when creating a new skill, editing an existing SKILL.md, or fixing one that misfires. Triggers: "์คํฌ ๋ง๋ค์ด์ค", "์ skill ์์ฑ", "SKILL.md ์จ์ค", "skill ๊ฐ์ ํด์ค", "create a skill", "skill documentation", "์คํฌ ๋ฌธ์ ์์ฑ", "workflow skill๋ก ๋ง๋ค์ด์ค".
Use when someone wants tests that prove a fix or feature actually works, not just tests that happen to pass โ TDD, red-green-refactor, falsifiable tests. Triggers: "TDD", "ํ ์คํธ ๋จผ์ ์์ฑ", "test-first", "red-green-refactor".
Use when a written implementation plan is about to be run โ critically review it, set each step's pass bar, then route to the right executor. Triggers: "๊ณํ ์คํํด์ค", "plan ์คํ", "์ด ๊ณํ ๊ฒํ ํ๊ณ ์งํ", "execute this plan", "plan ๋ฆฌ๋ทฐํ๊ณ ๊ฐ์", "์คํ ์ ์ ๊ธฐ์ค ์ก์์ค".
Use when facing 2+ independent jobs that can run concurrently โ fan out in parallel, each on its best-fit persona. Triggers: "๋ณ๋ ฌ๋ก ์ฒ๋ฆฌํด์ค", "๋์์ ์ฌ๋ฌ ์์ ", "parallel agents", "๋ ๋ฆฝ์ ์ธ ์์ ๋ค ํ๋ฒ์", "๊ฐ์ ์ ๋ฌธ๊ฐํํ ๋งก๊ฒจ์ค", "multiple independent failures".
Use before claiming work is done, tests pass, a build is green, or a bug is fixed without fresh isolated evidence. Triggers on: "์๋ฃํ์ด", "ํ ์คํธ ํต๊ณผ", "๋น๋ ์ฑ๊ณต", "๋ฒ๊ทธ ๊ณ ์ณค์ด", "์ปค๋ฐํ ๊ฒ", "PR ์ฌ๋ฆด๊ฒ", "done", "should pass", "bug is fixed".
Use when someone shares a developer portfolio and wants honest, interviewer-perspective feedback. Triggers on: "ํฌํธํด๋ฆฌ์ค ํผ๋๋ฐฑ ํด์ค", "๋ด ํฌํธํด๋ฆฌ์ค ์ด๋?", "portfolio review", "review my portfolio", "ํฌํธํด๋ฆฌ์ค ์ ์ ๋งค๊ฒจ์ค", "์ด๋ค ์ธํฐ๋ทฐ์ด๊ฐ ๋ณด๋ฉด ์ด๋ป๊ฒ ๋ณผ๊น?", "portfolio critique".
Use when reviewing or quality-checking a skill before shipping. Runs 6 checks: usefulness, authoring, structure, MCP, weight, output quality. Triggers on: ์คํฌ ๊ฒํ , skill review, validate skill, ๋ฐฐํฌ ์ ๊ฒํ .
Use when someone is overwhelmed by information, stuck in vague thinking, or spinning in analysis without moving forward. Three modes: cut signal from noise, eliminate vague language, and break overthinking loops. Triggers on: "overwhelmed", "too
Use when stress-testing a high-stakes decision, strategic plan, or argument end-to-end. Triggers on: "critical thinking full process", "decision stress-test", "์ ๋ต ๊ฒํ ์ ์ฒด", ๋ ผ๋ฆฌ ์ ์ฒด ์ ๊ฒ", "high-stakes decision review", "๊ฒฐ์ ์ ๋๋ก ๊ฒํ ํด์ค".
Use when someone is weighing options or making a decision where something is being given up. Makes the cost structure of each option explicit โ including hidden axes โ without picking a winner. Triggers on: "tradeoff", "ํธ๋ ์ด๋์คํ", "์ฅ๋จ์ ", "pros and
Use when someone needs to build a command-line tool โ defining subcommands, flags, and argument parsing; adding interactive prompts, progress bars, or shell completions; or distributing a cross-platform terminal application. Triggers on: "build a
Use when someone wants to look at their life from a distance โ not a specific decision, but the whole shape of it. Especially suited to life transitions, major milestones, or the feeling of running on autopilot. Triggers on: "์ถ์ ๋์๋ณด๊ณ ์ถ์ด", "๋ด ์ธ์์
Use when someone wants to talk, vent, or process emotions โ not solve a task. Counseling persona. Triggers: "์๊ธฐ ์ข ํ์", "์๋ดํด์ค", "์ค๋ ์ข ํ๋ค์ด", "talk to me", "I need to vent", "๋จธ๋ฆฌ๊ฐ ๋ณต์กํด", "๊ทธ๋ฅ ๋ค์ด์ค", "๊ธฐ๋ถ์ด ์ด์ํด".
Source text: Korean
Use when someone is unclear about what actually matters to them, feeling torn between competing priorities, or agonizing over a choice where something important has to be sacrificed. Two moves: extract genuinely held values from behavioral...
Use when a skill isn't triggering reliably on natural language or Korean input. Triggers on: "์คํฌ์ด ํธ๋ฆฌ๊ฑฐ ์ ๋ผ", "skill not firing", "description ๊ฐ์ ํด์ค", "ํ๊ตญ์ด ํธ๋ฆฌ๊ฑฐ ์ถ๊ฐํด์ค", trigger coverage audit", "skill ๋ฐ๋ ์กฐ๊ฑด ์ ๊ฒ", "description์ด ๋๋ฌด keyword๋ง ์์ด".
Use when capturing an architectural decision as an ADR โ after a Design Review or from raw context. Triggers: "ADR ์จ์ค", "์ด ๊ฒฐ์ ADR๋ก", "write an ADR", "document this decision". Fixed template; Decision must be declarative ("~๋ฅผ ์ฑํํ๋ค").
Use when something needs to be designed or built and the solution shape is unclear. Triggers on: "์ด๋ป๊ฒ ๋ง๋ค์ง?", "๊ธฐ๋ฅ ์ค๊ณํด์ค", "์ํคํ ์ฒ ์ก์์ค", "design this feature", "how should we build X?", "์ต์ ๋ ๋ฝ์์ค", "๋ฐ์ฐํด์ค". Always invoke before writing code.
Source text: Korean
Use when you need to think through a complex problem systematically โ generating ideas, stress-testing them, and converging on a decision. Triggers on: "think workflow", "๊น์ด ์๊ฒ ์๊ฐํด๋ณด์", "์์ด๋์ด๋ฅผ ์ฒด๊ณ์ ์ผ๋ก ๊ฒํ ", structured thinking", "์ฒด๊ณ์ ์ฌ๊ณ ํ๋ก์ธ์ค", "idea to
Source text: Mixed languages
Use when someone wants their idea, plan, or decision stress-tested with the strongest possible objections. Triggers on: "๋ฐ๋ก ๋์ ธ์ค", "์ฝ์ ์ฐพ์์ค", "์ด ๊ณํ ๋ฌธ์ ์ ์ด ๋ญ์ผ?", "devil's advocate", "๊ฒฐํจ ์ฐพ์์ค", "๋นํํด์ค", "์น๋ช ์ ์ฝ์ ", "steel-man", "punch holes in".
Source text: Mixed languages
Use when the current approach feels fundamentally wrong, inherited constraints need questioning, or someone wants to reason from the ground up. Triggers on: "์ ์ด๋ ๊ฒ ํด์ผ ํด?", "๊ธฐ๋ณธ๋ถํฐ ๋ค์ ์๊ฐํด", "first principles", "๊ฐ์ ์ ์์ฌํด๋ด", ์ด ๋ฐฉ์ ์์ฒด๊ฐ ๋ง๋ ๊ฑด์ง", "์ฒ์๋ถํฐ ๋ค์
Use when someone needs to negotiate, persuade, or navigate a hard conversation. Triggers on: "์ฐ๋ด ํ์", "salary negotiation", "๊ณ์ฝ ํ์", "๊ฑฐ์ ๋นํ์ด", "์ค๋ํด์ผ ํด", ํ์ ์ค๋น", "ask for a raise", "contract terms", "ํด๋ผ์ด์ธํธ ์ค๋", "์์ฌํํ ์ด๋ป๊ฒ ๋งํด".
Use when solutions keep feeling wrong or shallow โ when the problem itself may be mis-defined. Triggers on: "๋ญ๊ฐ ์๋ชป๋ ๊ฒ ๊ฐ์", "๊ณ์ ์ด ๋ฌธ์ ๊ฐ ๋ฐ๋ณต๋ผ", "๋ค๋ฅธ ๊ฐ๋๋ก ๋ด์ผ ํ ๊ฒ ๊ฐ์", "wrong problem", "solutions feel off", "์ด ๋ฌธ์ ์์ฒด๊ฐ ๋ง๋ ๊ฑด์ง", "์ฆ์๋ง ์๊พธ ์ก๋ ๋๋".
Source text: Korean
Use when raw, scattered ideas or notes need structure before writing, presenting, or deciding. Triggers on: "์๊ฐ ์ ๋ฆฌํด์ค", "๋ ธํธ ์ ๋ฆฌ", "๋จธ๋ฆฟ์์ด ๋ณต์กํด", "์์ด๋์ด ๊ตฌ์กฐํ", "outline ๋ง๋ค์ด์ค", mind map", "์ด๊ฑฐ ์ด๋ป๊ฒ ์ฎ์ด์ผ ํด?", "๊ธ ์ฐ๊ธฐ ์ ์ ๊ตฌ์กฐ ์ก์์ค", "์ฐ๋งํ ์๊ฐ ์ ๋ฆฌ".
Use when drafting a Design Review for a new feature or system. Triggers: "design review ์จ์ค", "๋์์ธ ๋ฆฌ๋ทฐ ๋ฌธ์", "RFC ์ด์", "write a design review", "์ด ๊ธฐ๋ฅ ์ค๊ณ ๊ฐ์ด ์ก์". Fixed 8-section template.
Use when someone needs to make or document architectural decisions from scratch โ choosing between system topologies (monolith vs. microservices), writing ADRs, evaluating scalability trade-offs, or selecting database and infrastructure patterns.
Use when designing a system from scratch, evaluating a monolith-to-MSA migration, or needing a structured end-to-end architecture process from domain discovery to documentation. Triggers on: "system architecture from scratch", "architecture
Use when code, an API, or a project lacks documentation and needs it created or improved โ adding docstrings or JSDoc to functions and classes, generating OpenAPI/ Swagger specs from an existing API, building a documentation site, or writing...
Plans and maintains a documentation system across architecture docs, API references, runbooks, and onboarding guides โ producing a doc coverage map and writing or improving the highest-leverage missing content. Use when a codebase, API, or team...
Use when someone is starting a new product, untangling a legacy system, mapping how a business process actually works, or asking where to begin domain modeling โ even if they do not name Event Storming explicitly. Triggers on: "domain modeling",