AgentSkillsCN

cv-optimization

当用户提出“优化我的简历”、“量身定制我的履历”、“为求职修改我的简历”、“让我的简历更符合 ATS 要求”、“根据职位描述优化我的简历”、“重写我的简历”，或需要简历优化、ATS 合规、关键词匹配，以及面试准备相关指导时，应使用此技能。

SKILL.md
--- frontmatter
name: cv-optimization
description: >
  This skill should be used when the user asks to "optimize my CV",
  "tailor my resume", "fix my CV for a job", "make my resume ATS-friendly",
  "improve my resume for a job description", "rewrite my CV", or needs
  guidance on resume optimization, ATS compliance, keyword matching,
  or interview-ready document preparation.
version: 0.1.0

CV / Resume Optimization

Optimize a CV or resume for a specific job description. The goal is to maximize interview callback rates while keeping the document ATS-friendly and human-sounding.

Core Principles

  1. •ATS compatibility first — Most resumes are filtered by Applicant Tracking Systems before a human ever sees them. Every optimization must pass ATS parsing.
  2. •Keyword alignment — Match the language, terminology, and phrasing used in the job description naturally throughout the CV.
  3. •Human readability — After passing ATS, a recruiter spends ~7 seconds scanning. The CV must be clear, scannable, and compelling to a human reader.
  4. •Authenticity — Never fabricate experience, skills, or qualifications. Reframe and emphasize existing experience to align with the target role.

Workflow

Step 1: Parse the Input CV

Read the uploaded CV (PDF, DOCX, or Markdown). Extract all content and identify sections dynamically. Common sections include but are not limited to:

  • •Contact Information / Header
  • •Professional Summary or Objective
  • •Work Experience / Employment History
  • •Skills (Technical and Soft)
  • •Education
  • •Certifications / Licenses
  • •Projects
  • •Volunteer Work
  • •Awards / Achievements
  • •Publications
  • •Languages

Preserve the original section structure unless restructuring is clearly beneficial.

Step 2: Analyze the Job Description

Parse the job description provided by the user. Extract:

  • •Required qualifications — hard requirements (years of experience, specific degrees, certifications)
  • •Preferred qualifications — nice-to-haves that provide an edge
  • •Key responsibilities — what the role actually does day-to-day
  • •Technical skills — tools, technologies, platforms, methodologies mentioned
  • •Soft skills — leadership, communication, collaboration keywords
  • •Industry terminology — domain-specific language and acronyms
  • •Action verbs — verbs the JD uses repeatedly (managed, led, developed, etc.)
  • •Company values / culture signals — anything about company mission or team culture

Step 3: Gap Analysis

Compare the CV against the JD:

  • •Strong matches — experience/skills that directly align with JD requirements
  • •Partial matches — related experience that can be reframed to better align
  • •Gaps — JD requirements not currently reflected in the CV
  • •Irrelevant content — CV content that adds no value for this specific role

For gaps, determine whether the user likely has the experience (just not highlighted) or genuinely lacks it. Never add fabricated experience.

Step 4: Section-by-Section Optimization

Optimize each section following the detailed guidelines in references/section-guidelines.md.

General rules for all sections:

  • •Use exact keywords from the JD where they truthfully apply
  • •Write in active voice with strong action verbs
  • •Quantify achievements wherever possible (numbers, percentages, dollar amounts, team sizes)
  • •Remove or de-emphasize content irrelevant to the target role
  • •Keep bullet points concise: 1-2 lines each, leading with the action/impact
  • •Avoid buzzwords that add no substance ("synergy", "guru", "ninja", "rockstar")
  • •Spell out acronyms on first use, then use the acronym (ATS systems search for both)

Step 5: ATS Compliance Check

Run through the ATS compliance checklist in references/ats-checklist.md to verify the optimized CV will parse correctly.

Step 6: Human Tone Review

Review the entire optimized CV for natural, human-sounding language:

  • •Does each bullet tell a mini-story? (Action → Context → Result)
  • •Would the user be comfortable saying these things in an interview?
  • •Is the language specific and concrete rather than vague and generic?
  • •Does it read like a professional wrote it, not a template generator?
  • •Are sentences varied in structure (not all starting the same way)?

Step 7: Output

Generate the final CV in the requested format (DOCX by default, PDF if requested).

Present a summary to the user:

  • •Key changes made per section
  • •Keywords incorporated from the JD
  • •Suggestions the user should verify (e.g., confirming specific metrics)
  • •Any gaps that could not be addressed without fabrication — suggest the user add relevant experience if they have it

Additional Resources

  • •references/section-guidelines.md — detailed optimization rules for each CV section
  • •references/ats-checklist.md — ATS compatibility verification checklist
  • •references/keyword-strategy.md — keyword identification and placement strategies