agentleFS
kyoungbinkimkyoungbinkim/give-me-job/skills/resume-intake/SKILL.md

resume-intake

Skill6 starsChanged 5 months agoMITView on GitHub

Build or improve a job seeker's resume.md source file from conversational career input. Use when the user wants to create resume.md, organize new-grad or experienced career history, convert raw experiences into STAR/CAR evidence, add missing metrics, or prepare reusable career evidence for cover letters and job applications. 한국어 요청도 포함합니다 - 이력서 작성, 이력서 정리, 경력기술서, 경력 정리, 커리어 정리, 자소서 소재 정리, 내 경험 정리해줘, 신입 이력서, 경력직 이력서.

The file

Show its rules and text

We split this skill into the rules it gives an agent

17 rules. About 0 are common, 9 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use this skill to turn raw career material into a structured resume.md that other give-me-job skills can trust.

    line 8
  2. Very specific:

    Use this skill when resume.md is missing, thin, outdated, or lacks evidence that can support a Korean cover letter.

    line 13
  3. Very specific:

    Do not use this skill when the user already has a usable resume.md and only needs JD analysis, cover letter drafting, HR review, or …

    line 17
  4. Distinctive:

    Follow user-provided facts only.

    line 23
  5. Distinctive:

    Do not upgrade a school project into professional work.

    line 32
  6. Distinctive:

    Do not turn participation into ownership unless the user says they owned the work.

    line 33
  7. Very specific:

    Read references/resume-schema.md before creating or rewriting resume.md.

    line 56
  8. Very specific:

    Read an existing resume.md if present.

    line 60
  9. Distinctive:

    Create or update the metadata block:

    line 63
  10. Distinctive:

    When the user states why they chose an approach, rejected an alternative, or changed direction, capture that reasoning too.

    line 87
  11. Very specific:

    Write Summary last. If any core work/project evidence remains Low, do not polish a final summary.

    line 96
  12. Distinctive:

    Keep claims factual. Do not invent company names, numbers, responsibilities, awards, links, tools, or outcomes.

    line 99
  13. Very specific:

    Do not create a separate Experience Bank for new intake.

    line 126
  14. Distinctive:

    Never copy a newly generated number or claim from a cover letter into the resume.

    line 136
  15. Distinctive:

    Treat missing contact fields and profile links as intake gaps.

    line 141
  16. Very specific:

    Use references/resume-schema.md for the canonical structure.

    line 154
  17. Distinctive:

    If the user gives too little information, ask for 3 short prompts:

    line 158

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Resume Intake

Use this skill to turn raw career material into a structured resume.md that other give-me-job skills can trust.

Trigger

Use this skill when resume.md is missing, thin, outdated, or lacks evidence that can support a Korean cover letter.

Do Not Trigger

Do not use this skill when the user already has a usable resume.md and only needs JD analysis, cover letter drafting, HR review, or packaging.

Autonomy Level

DoF: LOW

Follow user-provided facts only. Do not infer missing employers, dates, metrics, awards, tools, responsibilities, or outcomes.

Permitted inferences:

  • Career type from stated graduation date, work history, or years of experience.
  • Evidence strength from specificity, role clarity, and observable result.

Prohibited inferences:

  • Do not upgrade a school project into professional work.
  • Do not turn participation into ownership unless the user says they owned the work.

Input Contract

Required context:

  • Raw career material from the user or an existing resume.md.

Optional context:

  • Target role, target company, JD, or desired cover-letter questions.

Required parameters:

  • target_role: target role when available, otherwise unknown.
  • career_type: new-grad, experienced, or unknown.

Outputs produced:

  • resume.md draft, a patch/update plan, or focused follow-up questions.

Workflow

  1. Read references/resume-schema.md before creating or rewriting resume.md.
  2. Determine whether the user is new-grad, experienced, or still unknown.
  3. Read an existing resume.md if present. Preserve valid facts, but migrate old Profile, Core Summary, Experience Bank, or Work History structures into the canonical sections.
  4. Create or update the metadata block:
    • last-updated: today's date
    • career-type: new-grad or experienced
    • target-role: target role or Unknown
    • intake-version: increment the existing value, or use 1
  5. Collect raw experiences in the user's language first. Do not force every field before making progress, but ask for missing facts that materially affect evidence quality.
  6. Build the canonical top-level sections in this order:
    • Personal Info
    • Target Role
    • Summary
    • Work Experience for experienced candidates, or for new graduates only when they have relevant work/internship/part-time experience
    • Projects
    • Education
    • Skills
    • Certificates & Awards only when relevant
  7. Split work and project material into STAR/CAR bullets:

    • action verb
    • what was done
    • method, scale, or constraints
    • measurable or observable result

    When the user states why they chose an approach, rejected an alternative, or changed direction, capture that reasoning too. Cover-letter and interview questions frequently ask how a candidate decided, and that judgment cannot be recovered later from a result-only bullet. Ask for it when an entry is the candidate's strongest evidence. Never reconstruct a rationale the user did not give.

  8. End every work role and project entry with exactly one strength comment: <!-- strength: High -->, <!-- strength: Medium -->, or <!-- strength: Low -->.
  9. Write Summary last. If any core work/project evidence remains Low, do not polish a final summary; leave Pending stronger evidence. and ask follow-up questions.
  10. Keep claims factual. Do not invent company names, numbers, responsibilities, awards, links, tools, or outcomes.

New Grad Focus

Emphasize project depth, learning speed, role clarity, problem solving, collaboration, and job relevance. Projects is required even when work experience is absent. If metrics are weak, strengthen the explanation of process, decision making, and learning without pretending there was business impact.

New-grad results are usually small, so the reasoning behind them carries most of the signal. Prefer capturing how the candidate framed a problem, what they tried first, and what they changed after it did not work, over inflating the outcome. Also capture mistakes the candidate owns and what they changed afterward: 실패 경험 and 성장과정 questions are common, and an honest, resolved mistake is usable evidence that a polished result-only bullet cannot supply.

Experienced Focus

Separate responsibility from achievement. Work Experience is required for experienced candidates. Emphasize role scope, measurable result, business impact, cross-functional work, decision making, and repeatable contribution.

Schema Decisions

Adopt the canonical resume-style structure in references/resume-schema.md. Do not create a separate Experience Bank for new intake. Assign stable IDs such as EXP-001 to evidence in Work Experience or Projects, preserving existing IDs. Record the source section, entry title, and current line/bullet location alongside each ID. Never renumber IDs when entries move.

For each experience capture situation, the candidate's role, action, result, learning, and suitable question types. Keep company/question reuse and related follow-up questions as usage notes, separate from career facts. Rewrite reused evidence for the selected company, role, and question purpose.

Never copy a newly generated number or claim from a cover letter into the resume. When the user provides a new fact, show the proposed factual change before updating the resume and related materials. Changed source facts require revalidation of packages that used them.

Treat missing contact fields and profile links as intake gaps. Ask for them or mark them clearly as missing; do not fabricate placeholder URLs, phone numbers, schools, employers, or metrics.

Output

Return one of these:

  • a complete resume.md draft using the canonical schema
  • a patch/update plan for the existing resume.md
  • focused follow-up questions when evidence is too weak or required personal info is missing

Use references/resume-schema.md for the canonical structure.

Fallback

If the user gives too little information, ask for 3 short prompts:

  • target role
  • strongest project or work experience
  • one measurable or observable result

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.