菜单
Use when someone provides both a portfolio and a specific job description and wants to know how well they match — gap analysis, fit score, and positioning advice for that exact role. Triggers on: "이 공고에 맞아?", "JD랑 비교해줘", "이 포지션 지원해도 돼?", "match my
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/newkayak12/claude-skills --skill portfolio-jd命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Use when a substantial request needs verified, non-repetitive output. Triggers on: "이거 제대로 해줘", "검증까지 해서", "하네스 돌려줘", "run the harness", "do this properly with verification", "6단계로 처리해줘", "메타스크립트로". Not for trivial edits or Q&A.
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.
基于 SOC 职业分类
正在显示 SKILL.md
| name | portfolio-jd |
| description | Use when someone provides both a portfolio and a specific job description and wants to know how well they match — gap analysis, fit score, and positioning advice for that exact role. Triggers on: "이 공고에 맞아?", "JD랑 비교해줘", "이 포지션 지원해도 돼?", "match my |
| scenarios | ["Compare my portfolio to this job description — where are the gaps?","I want to apply to this JD — how well does my portfolio match and what should I fix?","Score my portfolio against this job posting across tech stack, scale, and role scope","이 공고에 내 포트폴리오가 맞는지 갭 분석해줘","이 JD랑 포트폴리오 비교해서 서류 통과 가능성 알려줘"] |
| compatibility | {"recommended":["think-tool","mcp-reasoner"],"optional":["sequential-thinking"],"remote_mcp_note":"think-tool은 JD 파싱(Stage 1)과 포트폴리오 파싱(Stage 2)에서 필수 체크포인트로 사용됩니다. mcp-reasoner는 갭 심각도 분류(치명/보완 가능/마이너)와 서류 통과 가능성 판단에 사용됩니다 — 이 두 가지가 이 스킬의 가장 중요한 판단입니다. Claude 설정 → MCP Servers에서 remote SSE 엔드포인트를 추가하세요."} |
Use when:
Not for:
Provide:
| Claude | You |
|---|---|
| Decodes JD: real requirements vs. aspirational nice-to-haves | Provides the actual portfolio content |
| Scores fit across 5 dimensions with evidence | Validates gap assessment with network contacts if possible |
| Classifies each gap: critical / recoverable / minor | Decides whether to apply and what to change |
| Gives concrete portfolio positioning advice for this specific role | Makes actual application decision |
| Gives honest pass/borderline/screen-out assessment | Does the relationship building to get referrals |
../portfolio-rewrite/SKILL.md — act on specific gap areas identified here../resume-tailorer/SKILL.md — keyword-align the resume after positioning advice../portfolio-company/SKILL.md — company type fit analysis if no specific JD yetCall think before reading the portfolio. Build the JD profile independently first — comparing too early biases the analysis toward the portfolio's framing.
Required questions:
🧠 JD note: Record decoded profile here. "The role is actually looking for X, not just Y as stated."
Call think to build the candidate profile independently of the JD. Do not reference the JD yet.
Required questions:
🧠 Portfolio note: Record candidate profile here. This is your independent read before the comparison.
Call sequentialthinking to plan how to map portfolio signals to JD requirements systematically:
Evaluate across these dimensions:
기술 스택 매칭 Does the candidate's tech experience overlap with what the JD requires? Note exact matches, adjacent experience (transferable), and gaps.
경험 연차 / 스케일 매칭 Does the candidate's scale of experience match what the role implies? A JD saying "large-scale systems experience preferred" is telling you something.
역할 범위 매칭 Does the JD expect an IC who executes, or a senior who leads and influences? Does the portfolio show the right kind of contribution?
도메인 매칭 Is there relevant domain experience (fintech, e-commerce, infra, data platforms, etc.)? Domain mismatch is often recoverable, but it should be acknowledged.
소프트 시그널 매칭 JDs often encode cultural expectations. "Proactive", "ownership", "self-directed" are different from "collaborative", "process-oriented", "strong communicator." Does the portfolio's writing style match the cultural signal?
Write in Korean. Use this structure:
[역할 해석] What is this role actually looking for, beyond what the JD literally says? 3–5 sentences.
[종합 매칭 점수] X / 10 — one sentence explanation of the score.
Break down as:
Insert 🧠 JD note where any score required a non-obvious judgment.
[강한 매칭 포인트] What in the portfolio directly speaks to this JD. Quote or reference specific portfolio content and the specific JD requirement it addresses.
[갭 분석] Where the portfolio falls short for this specific role.
For each gap, call mcp-reasoner (beam_search, beamWidth=3) before assigning severity:
For each gap:
[포트폴리오 포지셔닝 조정] If the candidate applies, how should they frame their experience for this specific role? What to emphasize, what to downplay, what to add to the portfolio or cover letter before applying.
Concrete: "이 JD는 데이터 파이프라인 경험을 중요하게 보는데, 포트폴리오에서 [프로젝트명]의 ETL 작업을 현재보다 더 앞에 배치하고, 처리한 데이터 볼륨을 명시하세요."
[서류 통과 가능성 평가]
Call mcp-reasoner (mcts, numSimulations=50) for this judgment — it is the highest-stakes single output of this skill and the one most likely to have competing evidence.
Honest assessment: likely pass / borderline / likely screen out — and what that hinges on.
State the one factor that, if changed, would most shift this assessment in either direction.
[지원 여부 조언] Should they apply? If yes, with what changes? If borderline, what's the one thing that would tip it?
think whenever a match or gap is ambiguous — adjacent experience, partial domain overlap, or transferable skills all need careful judgment.