Tailor Dutch SWE CV

Use this skill to turn a target Dutch software engineering vacancy plus Svetlana's current CV into an ATS-safe, human-persuasive application package.

Workflow

1. Load references/profile-and-playbook.md for durable profile facts and the Dutch SWE screening model.
2. Inspect the current CV. Prefer the user-provided file; otherwise look for /Users/svetlana/Desktop/Senior_SWE_Perekrestova_CV.pdf and parse locally. Do not upload the CV to third-party parsing services.
3. Fetch the target job posting. If the user provides a URL, browse/scrape the current page and cite it. Identify ATS/portal, role title, location, office policy, required questions, and required/nice-to-have keywords.
4. Build a hard-filter table: location, office/hybrid, HSM/visa transfer, sponsorship, English/Dutch, years, stack, domain, screening/VOG/PES, salary/IND feasibility if relevant.
5. Build a keyword matrix using exact vacancy terms, grouped into technologies, domain, delivery practices, and culture/process.
6. Compare the CV against the vacancy. Flag mismatches, missing keywords, ATS parse risks, unclear eligibility, and human-review doubts.
7. Rewrite the CV around truthful fit. Keep one-column plain text structure, visible LinkedIn/GitHub URLs, exact keywords in context, quantified bullets, and no sensitive unnecessary data.
8. Draft application answers for required fields, especially motivation, office availability, work authorization/sponsorship, and recruiter-screen talking points.
9. Save deliverables in the current workspace outputs/ folder:
   • <company>-<jobid>-cv-screening-analysis.md
   • <company>-<jobid>-final-cv.md
   • <company>-<jobid>-final-greenhouse-answers.md or portal-specific equivalent
   • a DOCX CV when useful for ATS submission.
10. Verify no placeholders remain in final files and that the key target terms and eligibility facts are present.

Defaults

• Use saved facts from the reference file as defaults only when they do not conflict with the current CV or user message.
• Ask concise follow-up questions only for facts that materially affect submission, such as visa status, office availability, salary floor, language level, or whether a specific technology claim is truthful.
• If the job is not Dutch-market SWE, adapt the workflow but keep the ATS and human-review discipline.

Output Shape

For each target role, produce:

• concise verdict: strong / viable with risks / weak fit
• hard-filter analysis
• ATS keyword matrix
• current CV risk report
• rewritten top section, skills, and experience bullets
• final submit-ready CV text or DOCX
• application answer drafts
• final pre-submit checklist

Avoid

• Do not keyword-stuff or add hidden text.
• Do not invent experience, screening clearance, language level, degree equivalence, or work authorization.
• Do not preserve AI/LLM framing when the role is primarily fintech/platform/backend unless the vacancy values it.
• Do not include BSN, passport number, date of birth, marital status, full home address, health details, or family status.