splash0047/resume-generator / skills/resume-ats-optimizer
Resume ATS Optimizer
VerifiedValidate resume parsing, map job requirements to evidence, and improve role-specific machine readability without inventing universal ATS scores.
Resume ATS Optimizer
When to Use This Skill
Use this skill when the user wants to:
- optimize a resume for a specific job;
- diagnose low response rates;
- check whether a PDF/DOCX is likely to parse cleanly;
- identify missing required qualifications or weak evidence;
- compare a resume against a job description.
Do not present an internal heuristic as the employer's ATS score or claim that a particular percentage guarantees progression.
Core Principle
Treat resume screening as a sequence of distinct checks:
- Eligibility — graduation batch, degree, location/work authorization, availability, required years of experience, and other hard application conditions.
- Parsing — can the submitted file be extracted in a sensible reading order?
- Requirement evidence — does the resume show credible evidence for the job's required and preferred criteria?
- Human readability — can a recruiter quickly identify role fit, recent education/experience, and relevant projects?
- Downstream stages — assessments and interviews are separate from resume quality and must be diagnosed separately.
Different employers configure different workflows. Never imply that all ATS products rank candidates using keyword density.
Step 1 — Eligibility Gate
Extract explicit conditions from the job description and application form. Classify each as:
- Hard requirement — missing it may make the application ineligible.
- Preferred qualification — useful but not automatically disqualifying.
- Unknown / needs confirmation — wording is ambiguous.
Examples: expected graduation year, student status, degree discipline, location, work authorization, internship dates, minimum professional experience.
If a hard requirement is not met, state that clearly before optimizing wording.
Step 2 — Parsing Validation
For the actual exported file:
- verify it is text-based rather than scanned;
- extract the text and inspect reading order;
- confirm name, contact details, education, dates, skills, links, and section headings survive extraction;
- check whether application fields populated from the upload are correct;
- avoid graphics, photos, text boxes, complex tables, multi-column layouts, and contact details placed only in headers/footers;
- use the employer's requested file format.
A parser warning is not automatically a rejection. Report exactly what failed or was lost.
Step 3 — Requirement-to-Evidence Matrix
Build a table with these columns:
| Job criterion | Type | Evidence in resume | Evidence strength | Action |
|---|---|---|---|---|
| Python | Required | Root Cause Analyzer + internship | Direct | Keep and lead with it |
| AWS | Preferred | AWS re/Start credential only | Indirect | Do not imply production AWS ownership |
| 2+ years experience | Required | Student internships only | Missing | Eligibility risk |
Evidence strength:
- Direct — implemented, used, measured, or owned in a project/experience with supporting detail.
- Indirect — coursework, certification, README claim, or adjacent experience.
- Missing — no truthful evidence available.
Do not add a keyword unless it accurately describes the candidate's work or credential.
Step 4 — Role-Specific Keyword Use
Use exact job terminology when truthful, but do not optimize for repetition counts.
Priority:
- Technical Skills for concise discoverability.
- Experience bullets for proof of use.
- Project bullets for engineering depth.
A required term appearing once with strong evidence is more useful than repeating it several times without proof.
Step 5 — Resume Output Rules
- Use standard section names: Education, Experience, Projects, Technical Skills.
- For students/new graduates, make expected graduation explicit.
- Keep the resume focused on one role family at a time.
- Prefer 2–3 relevant projects over a long inventory.
- Use measured outcomes when available; otherwise use concrete implementation detail.
- Do not fabricate percentages, user counts, latency improvements, production scale, or "production-ready" claims.
- Distinguish training programs/badges from exam-based certifications.
- Ensure linked repositories visibly support project claims.
Diagnostic Output
Return:
- Eligibility findings
- Parsing findings
- Requirement-to-evidence matrix
- Unsupported or ambiguous claims
- Recommended role-specific edits
- Export validation checklist
- Outcome-tracking recommendation
You may compute a private coverage ratio such as "7 of 9 required criteria have direct evidence" for explanation. Do not label it a universal ATS score or assign a pass/fail threshold unless the employer publishes one.
Outcome Tracking
When the user is applying repeatedly, track:
- job URL and role family;
- eligibility status;
- resume version;
- application channel;
- submission date;
- outcome: pending, rejected before assessment, assessment received, interview received, offer/reject after interview.
Use those outcomes to diagnose where progression stops. Do not assume every rejection is caused by ATS parsing or keywords.
Guardrails
- Never promise a shortlist, callback rate, or interview multiplier.
- Never invent a metric to make a bullet look stronger.
- Never equate a parser issue with automatic rejection unless the employer documents that behavior.
- Never claim a certification proves hands-on production experience.
- Never hide an eligibility mismatch with wording.
- Always prefer verifiable evidence over keyword density.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |