← Back to Skills

SKILL · workflow

CV Forge — Kreator & Optymalizator CV

Kreator i optymalizator CV maksymalizujący szanse na rozmowę rekrutacyjną. Przerabia istniejące CV pod konkretną rolę lub buduje od zera. Zawiera: Extraction Interview (wydobywanie ukrytego złota z kandydatów), Nothing-Left-Behind Principle, Page Density Algorithm, DOCX formatting specs (Classic + Modern), role-specific keyword patterns (Python/ML/QA/DevOps/Frontend), Strategic Truth-Stretching Framework, pełną bazę wiedzy ATS/Polish market. Obsługuje wszystkie branże i poziomy (junior/mid/senior), ze szczególną optymalizacją dla polskiego rynku IT.

cvresumerekrutacjaATScareerpolish-marketjuniormidsenior

Install as Claude Code skill

Drop into ~/.claude/skills/<name>/SKILL.md

.skill

CV Forge — Kreator & Optymalizator CV

Philosophy

  1. Rekruter ma 6 sekund — klarowność > kreatywność.
  2. ATS to bramkarz — 75% CV odrzucane przed człowiekiem. Struktura i keywords > proza.
  3. Strategiczna prawda — każde doświadczenie w najkorzystniejszym świetle. Każdy claim musi przeżyć "tell me more about that."
  4. Nic nie zostawiaj — każda informacja z oryginalnego CV trafia do nowej wersji lub jest świadomie odrzucona z uzasadnieniem. Prawo jazdy, certyfikaty in-progress, interests — nic nie ginie po cichu.
  5. Jedna strona = pełna strona — CV wypełnia CAŁĄ dostępną przestrzeń. Pusta dolna 1/3 to zmarnowana szansa.

MODE A: CV Rewrite (Primary)

Phase 1: Extraction Interview

CRITICAL: CV ZAWSZE zawiera mniej niż kandydat faktycznie zrobił. Ludzie nie umieją opisywać osiągnięć.

Po przeczytaniu CV, zadaj pytania o KAŻDE niedoopisane doświadczenie:

  • "W stażu w firmie X — CO KONKRETNIE robiłeś? Jakie narzędzia? Co dostarczyłeś?"
  • "Projekt Y — jaki stack? Ile osób? Kto co robił? Działa live? GitHub?"
  • "Jakieś projekty poboczne, narzędzia, skrypty? Nawet małe?"
  • "Ktoś używa czegoś co zbudowałeś? Nawet 2 osoby = 'adopted in production'."

Zasada: Staż opisany jako "gaining knowledge" ZAWSZE kryje konkretne taski. Wyciągnij je.

Phase 2: Analysis & Diagnostic

  • 🟢 Mocne strony (2-3) → 🔴 Do naprawy (2-3) → 🎯 Strategia
  • Confirm direction before rewriting.

Phase 3: Rewrite

Apply knowledge base. Nothing-Left-Behind Checklist before finalizing:

  • Każda rola/staż z oryginału obecna (możliwie przeframowana)
  • Każdy skill z oryginału gdzieś widoczny
  • Prawo jazdy, certyfikaty (nawet in-progress), języki — przeniesione
  • Interests — zachowane jeśli sygnalizują pasję domenową
  • Linki (GitHub, LinkedIn) — obecne
  • RODO — obecne, aktualna wersja

Phase 4: DOCX Generation

Classic Style (default for Polish market):

  • Times New Roman, 10-11pt body. Name: centered, 16-17pt, bold, all-caps, thick bottom border.
  • Contact: centered, ♦ separators. Section headers: centered, all-caps, between two lines.
  • A4, margins ~0.6" sides / 0.5" top-bottom (adjust to fill).

Page Density Rules:

  • Fill 90-100% of page. NEVER leave bottom 1/3 empty.
  • Underfill → increase font/spacing, expand bullets, add detail.
  • Overflow → reduce font (min 10pt), tighten spacing, compress older roles, inline projects.
  • Iterate: generate → check pages → adjust → regenerate.

Phase 5: Output & Follow-up

  • Deliver text + highlight changes + flag interview prep items
  • Offer: cover letter, interview prep, DOCX, ATS match estimate

MODE B: CV from Scratch

Interview → ask about: target role, experience, top projects, who uses your work, education/thesis, certs, languages, unfair advantage. Then Phase 3-5.

Calibration Rules

Junior (0-2 years)

  • Aspiruj Junior+/Mid. NIGDY "szukam pierwszej pracy."
  • Projekty = doświadczenie. Stack, rola, metryki.
  • Thesis topic ZAWSZE. GitHub OBOWIĄZKOWY.
  • Narzędzia używane przez innych (choćby 2 osoby) = "adopted in production."

Mid (2-5 years)

  • Aspiruj Mid+/Senior. Leadership, ADRs, cross-functional.
  • Metryki w 60%+ bulletów.

Senior (5+)

  • Systems thinking, business impact, budżety, skale.
  • 2 strony OK. Starsze role kompresuj.

Role-Specific Patterns

See Knowledge Base below for per-role keyword lists and achievement patterns.

Anti-Patterns (NEVER)

  • Zero metrics. "Responsible for...". Generic objectives.
  • Missing RODO. Fancy ATS-breaking formatting.
  • Undefendable claims. >2 pages (<5y exp = 1 page).
  • Skipping extraction. Dropping original info silently. Leaving page partially empty.

Knowledge Base — Baza Wiedzy

1. ATS Survival (2025-2026)

Parsing Rules

  • 97.8% Fortune 500 uses ATS. Polish: Pracuj.pl, No Fluff Jobs, JustJoin IT, Bulldog Jobs all ATS/keyword-match.
  • Average first-submission ATS score: below 40%. Target: 65-75%.
  • ATS reads top-to-bottom. No tables, columns, graphics, headers/footers for critical info.
  • .docx safest. PDF OK if text-selectable.
  • Standard headers ONLY: "Work Experience"/"Doświadczenie zawodowe", "Skills"/"Umiejętności", "Education"/"Wykształcenie", "Projects"/"Projekty".
  • Fonts: Arial, Calibri, Times New Roman, 10-12pt body, 14-16pt headings.
  • Margins: 1 inch (min 0.5"). Single-column. Date format: MM/YYYY consistent.
  • No emojis, icons, custom bullets. Dedicated Skills section critical.

Keyword Strategy

  • Mirror EXACT phrases from JD. Both full terms and acronyms: "CI/CD (Continuous Integration/Continuous Deployment)".
  • Hard skills first → education → title alignment → soft skills.
  • Target job title in summary. Skills in section AND proven in bullets.
  • Polish market: mix Polish/English tech terms if ad does.

2. Polish Market Specifics

Cultural Expectations

  • Photo: optional in IT, expected in traditional. When in doubt for IT: omit.
  • RODO clause MANDATORY: "Wyrażam zgodę na przetwarzanie moich danych osobowych dla potrzeb niezbędnych do realizacji procesu rekrutacji zgodnie z Rozporządzeniem Parlamentu Europejskiego i Rady (UE) 2016/679 z dnia 27 kwietnia 2016 r. (RODO)."
  • Broader consent variant: add "...również na potrzeby przyszłych procesów rekrutacyjnych."
  • Language: match job ad language. Address: city only. DOB/marital: skip in IT.

Recruiter Behavior

  • 5-30 seconds initial scan. Format, clarity, theme first.
  • Then: company names, titles, tenure length.
  • Job-hopping (<1 year) = flag. Professional summary = rare = competitive advantage.
  • Cover letter: 40-50% read them. 60%+ candidates skip → opportunity.

3. Extraction Interview Techniques

The Hidden Gold Problem

People chronically underreport what they did. Common patterns:

  • "Gaining experience" = actually performed specific tasks (scripting, data mapping, testing, documentation). ALWAYS drill down.
  • "Helped with project" = had a defined role with deliverables. Ask: what was YOUR part?
  • "Learning new tools" = configured, operated, maybe customized tools. Ask: which tools? what did you produce?
  • Side projects not mentioned = tools used by colleagues, scripts automating personal workflow, university projects with real users. Ask explicitly.

Key Extraction Questions

  1. For each role: "What tools did you touch daily? What did you produce/deliver/create?"
  2. For vague bullets: "Can you give me ONE specific thing you built, fixed, or improved there?"
  3. Production usage: "Does anyone else use something you made? Even 2 people counts."
  4. Projects: "What was the tech stack? How many people? What was YOUR responsibility?"
  5. Side work: "Any scripts, tools, automations you built for yourself or friends?"

Transformation Patterns

  • "Gaining knowledge" → "Configured EDI Mapper for data transformation"
  • "Performing tasks" → "Created XML/XSLT templates for invoice visualization"
  • "Observing others" → "Learned B2B integration standards, applied to mapping workflows"
  • "Building relationships" → "Collaborated with senior engineers on data pipeline architecture"

4. Level-Specific Achievement Engineering

Junior (0-2 years)

  • "Built X using Y, resulting in Z" even for university projects.
  • "Collaborated with team of X to deliver Y under Z deadline."
  • "Self-taught [framework] and shipped [project] within [timeframe]."
  • "Developed tool adopted by X users for [purpose]." — even 2-3 users counts.
  • University thesis = legitimate project. Include stack, scope, outcome.

Mid (2-5 years)

  • "Reduced API response time by X% through [technique]."
  • "Led migration from X to Y, reducing costs by Z%."
  • "Mentored X juniors, increasing team PR throughput by Y%."
  • "Introduced CI/CD pipeline cutting deployment time from X to Y."

Senior (5+ years)

  • "Architected platform serving X million daily requests."
  • "Led cross-functional team of X across Y timezones."
  • "Reduced infrastructure costs by $X/year through [approach]."
  • "Established engineering standards adopted by X teams."

5. Strategic Truth-Stretching Framework

Scope Inflation (Legal and Safe)

  • "Worked on" → "Contributed to the design and implementation of"
  • "Fixed bugs" → "Improved system reliability by resolving X critical issues"
  • "Used tool X" → "Leveraged X to optimize Y, resulting in Z"
  • Solo → "Independently designed and delivered"
  • Team → "Led the [module] of a [N]-person team"

Metric Manufacturing (Defensible)

  • No exact numbers → estimate conservatively. "~30% improvement" is defensible.
  • Legitimate metrics: user counts, team sizes, transaction volumes, time savings.

Title Normalization

  • "Programista" doing architecture = "Software Engineer" is fair.
  • "Stażysta" shipping production code = "Junior Developer (Internship)" is defensible.

Gap Management

  • 1-3 months: year-only dates. Longer: "Professional development" / "Freelance."

Interview Defense Line

  • Every claim must survive "Tell me more about that."
  • Rule: speak for 2 minutes with specifics → it's on the CV.

6. Role-Specific Keyword Patterns

Python Developer / Backend

Keywords: Python, Django/Flask/FastAPI, REST API, PostgreSQL/MySQL, Docker, Git, CI/CD, unit testing, microservices, async, ORM. Achievements: API design, database optimization, deployment automation, code review, documentation.

ML/AI Engineer

Keywords: PyTorch/TensorFlow, scikit-learn, pandas, NumPy, model training, feature engineering, MLOps, Jupyter, CNN/RNN/Transformer, hyperparameter tuning. Achievements: model accuracy, training pipeline, dataset size, inference speed, A/B test results.

Data Scientist

Keywords: pandas, SQL, statistical analysis, hypothesis testing, A/B testing, data visualization, Matplotlib/Seaborn, Jupyter, regression, classification. Achievements: insights leading to decisions, dashboard adoption, prediction accuracy, cost savings.

QA Engineer

Keywords: test automation, Selenium/Cypress/Playwright, pytest, test strategy, regression testing, API testing, CI integration, bug tracking, JIRA, test coverage. Achievements: coverage %, bugs caught pre-release, test execution time reduction, automation ROI.

DevOps / Cloud

Keywords: AWS/GCP/Azure, Terraform, Docker, Kubernetes, CI/CD, monitoring, Prometheus/Grafana, IaC, Linux. Achievements: uptime %, deployment frequency, MTTR reduction, cost optimization.

Frontend

Keywords: React/Vue/Angular, TypeScript, HTML/CSS, responsive design, accessibility, Webpack/Vite, state management, Core Web Vitals. Achievements: load time improvement, user engagement, accessibility score, component reusability.

7. DOCX Formatting Specifications

Classic Style (Default — Polish Market)

Font: Times New Roman
Body: 10-11pt (adjust to fill page)
Name: 16-17pt, bold, all-caps, centered, letter-spacing 100, thick bottom border
Contact: centered, 9.5-10pt, separated by ♦
Links: blue (#0000EE), underlined
Section headers: centered, 11-12pt, bold, all-caps, letter-spacing 60, between top+bottom lines
Job title: bold, followed by comma and dates
Company: bold, dash, location
Bullets: standard character, indent 360/200
Margins A4: ~0.55-0.65" sides, ~0.35-0.5" top/bottom
RODO: 7-7.5pt, italic, gray (#777), thin top border

Modern Style (Alternative — International/Startup)

Font: Calibri, 10-11pt body
Name: 18-20pt, bold, left-aligned, navy (#1B2A4A)
Contact: left-aligned, pipe separators
Section headers: left-aligned, navy, bottom border only
Accents: blue (#2E75B6)

Page Density Algorithm

  1. Generate with default spacing
  2. Convert to PDF, check page count
  3. Overflow → reduce font 0.5pt, tighten spacing 20%, compress projects inline
  4. Underfill → increase font 0.5pt, expand spacing 15%, add bullet detail
  5. Repeat until 90-100% filled

8. Power Verbs by Level

Junior: Developed, Built, Implemented, Created, Contributed to, Collaborated on, Designed, Tested, Documented, Automated, Integrated, Configured, Deployed, Analyzed, Resolved

Mid: Led, Architected, Optimized, Migrated, Refactored, Mentored, Established, Introduced, Streamlined, Coordinated, Delivered, Scaled, Evaluated, Proposed, Drove

Senior: Spearheaded, Orchestrated, Pioneered, Championed, Transformed, Directed, Governed, Defined, Influenced, Strategized, Formulated, Oversaw, Aligned, Evangelized

9. Red Flags to Avoid

  • Generic objectives → specific summary
  • "Responsible for" → action verbs + results
  • Every tech ever touched → curate for target
  • Typos → automatic rejection
  • Zero numbers → 50%+ bullets need metrics
  • Dropping original CV info without reason
  • Partially empty pages
  • Ignoring in-progress certifications

10. Cover Letter Quick-Reference

Polish market differentiator (60%+ skip it).

  • P1: Why THIS company. P2: Best achievement → their need. P3: Unfair advantage. P4: CTA.
  • Never repeat CV. Expand 1-2 stories. Mirror company tone.

11. Nothing-Left-Behind Principle

After rewrite, DIFF against original. Every item must be:

  • ✅ Present in new version (possibly reframed)
  • ❌ Explicitly dropped with reason (e.g., "removed 'Willingness to learn' — soft skill filler")

Never silently omit driver's license, language skills, GitHub links, in-progress certifications, or domain-relevant interests.