resume-jd-tune
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseResume JD Tune: Job-Description Keyword Tuner
Resume JD Tune：职位描述关键词调整工具
Invocation:
/resume tune <resume-file> <jd-file-or-url>Scripts: Inline (no script needed; runs in the orchestrator).
Tailors an existing ATS-safe resume to a specific job description by mirroring the JD's hard-skill keywords and bridging title vocabulary. Targets Jobscan's 75–80% match benchmark.
调用方式：
/resume tune <简历文件> <JD文件或链接>脚本： 内置（无需额外脚本；在编排器中运行）。
通过匹配JD中的硬技能关键词和过渡头衔词汇，将现有符合ATS标准的简历调整为适配特定职位描述的版本，目标是达到Jobscan的75%-80%匹配基准。
Behavior
行为规则
Always interview the user briefly before tuning. Use (one question per call) to ask:
mcp__conductor__AskUserQuestion- Confirm target role title from JD vs. user's actual seniority claim
- For each JD hard-skill the resume doesn't already mention — does the user actually have this skill? (Yes / No / Limited exposure)
- Tone preference for any rewritten bullets (Conservative / Modern / Balanced)
- Whether to add a bridging title line if titles differ materially
Never claim a skill the user marks "No". Flag the gap in the report instead.
调整前务必先与用户进行简短沟通。 使用（每次调用仅提出一个问题）询问以下内容：
mcp__conductor__AskUserQuestion- 确认JD中的目标职位头衔与用户实际的资历声明是否一致
- 对于简历中未提及的每个JD硬技能——用户是否确实掌握该技能？（是/否/有限接触）
- 重写项目符号内容的语气偏好（保守/现代/平衡）
- 若当前头衔与目标头衔存在较大差异，是否添加过渡头衔行
绝不要将用户标记为“否”的技能添加到简历中，而是在报告中注明该技能缺口。
Quick Reference
快速参考
| Step | What it does |
|---|---|
| 1 | Brief interview (skill confirmation, tone, bridging-title preference) via AskUserQuestion. |
| 2 | Parse the JD — extract hard skills, required tools, exact title, key phrases. |
| 3 | Diff against the resume's Skills section + most-recent-role bullets. |
| 4 | Insert missing JD keywords into Skills (only if user confirmed they have the skill). |
| 5 | Rewrite 1–2 bullets in the most-recent role to use JD's exact phrasing. |
| 6 | Add bridging title line under the name if the user's title differs from the JD title. |
| 7 | Report projected match score before/after, plus skill gaps. |
| 步骤 | 操作内容 |
|---|---|
| 1 | 通过AskUserQuestion进行简短沟通（技能确认、语气偏好、过渡头衔意愿）。 |
| 2 | 解析JD——提取硬技能、必备工具、准确职位头衔、关键表述。 |
| 3 | 对比简历的技能板块和最新任职经历的项目符号内容。 |
| 4 | 将缺失的JD关键词添加到技能板块（仅当用户确认掌握该技能时）。 |
| 5 | 重写最新任职经历中的1-2个项目符号，使用JD中的准确表述。 |
| 6 | 若用户当前头衔与JD中的头衔不同，在姓名下方添加过渡头衔行。 |
| 7 | 报告调整前后的预计匹配分数，以及技能缺口。 |
Rules
规则说明
- Never claim a skill the user doesn't have. If the JD lists "Kafka" and the user has never touched it, do NOT add it. Flag the gap to the user.
- Mirror exact phrasing where possible. If the JD says "Kubernetes," include both "Kubernetes" and "K8s" in Skills (one parses each variant).
- Acronym + expansion once. "SEO (Search Engine Optimization)" the first time only.
- Bridging title is optional. Only add if the user's title is materially different from the JD's. "Senior Software Engineer" → "Senior Software Engineer | Backend Platform Engineer" is fine.
- Don't keyword-stuff. AI-layered ATS in 2026 (Jobscan AI, Eightfold) cross-check claims against bullet evidence and penalize unsupported skills.
- 绝不要虚报用户未掌握的技能。若JD中列出“Kafka”但用户从未接触过，请勿添加该技能，而是向用户注明此缺口。
- 尽可能匹配准确表述。若JD中写的是“Kubernetes”，请在技能板块同时包含“Kubernetes”和“K8s”（两种变体均可被解析）。
- 首字母缩写+全称仅出现一次。首次出现时使用“SEO (Search Engine Optimization)”即可。
- 过渡头衔为可选内容。仅当用户当前头衔与JD中的头衔存在较大差异时才添加。例如“高级软件工程师”→“高级软件工程师 | 后端平台工程师”是合理的。
- 不要堆砌关键词。2026年的AI驱动ATS系统（如Jobscan AI、Eightfold）会交叉验证技能声明与项目符号中的证据，无支撑的技能会被扣分。
Output
输出内容
A modified DOCX named , plus a side-by-side report:
<original-name>_tuned_for_<jd-slug>.docxJD: Senior Backend Engineer @ AcmeCo
Match score: 72% → 87%
Hard skills: 14/19 → 19/19 (added: gRPC, Postgres, Kubernetes)
Title bridge: added "| Backend Platform Engineer"
Bullets edited: 2 (Aerem, Sky2c)
Skill gaps: [none — user had all JD skills]If there are skill gaps, the report calls them out so the user can decide whether to add evidence in a different format (project, certification, side work).
生成一个名为的修改版DOCX文件，以及一份对比报告：
<原文件名>_tuned_for_<jd-slug>.docxJD: Senior Backend Engineer @ AcmeCo
匹配分数: 72% → 87%
硬技能: 14/19 → 19/19 （新增：gRPC, Postgres, Kubernetes）
过渡头衔: 添加了"| Backend Platform Engineer"
修改的项目符号: 2项（Aerem, Sky2c）
技能缺口: [无——用户掌握所有JD要求的技能]若存在技能缺口，报告中会明确指出，方便用户决定是否通过其他形式补充相关证明（如项目经验、证书、副业成果）。
Sources
参考来源
Match-score benchmarks from Jobscan, Resume Worded, and Enhancv (2026). Methodology: keyword overlap weighted by section (Skills 1.0x, recent-role bullets 0.6x, summary 0.4x, older bullets 0.2x).
匹配分数基准来自Jobscan、Resume Worded和Enhancv（2026年数据）。计算方法：关键词重叠度按板块加权（技能板块1.0倍、最新任职经历项目符号0.6倍、个人简介0.4倍、过往任职经历项目符号0.2倍）。