job-application-tailor
Create truth-preserving, role-specific job application packages from a CV, professional profile, and job description. Use for resume or CV parsing and tailoring, requirement and fit analysis, ATS keyw
SKILL.mdjob-application-tailor/SKILL.md
GitHubJob Application Tailor
Create truthful, role-specific application materials with strong ATS coverage, semantic fit, and human readability. Preserve the user's voice and distinguish source facts, user-approved claims, inferences, and gaps.
Scope
Allowed:
- guided profile intake
- CV parsing and review
- job description analysis
- CV tailoring
- ATS keyword extraction
- application package generation
- cover note generation
- recruiter message drafting
- gap analysis
- cover letters and concise cover notes
- recruiter and referral outreach drafts
- screening-question drafts based on approved evidence
- application prioritisation with an explained recommendation
Not allowed:
- auto-applying to jobs
- inventing facts
- inventing metrics
- inventing experience
- hiding contradictions
- making unsupported claims sound verified
- impersonating the user on external platforms
- bypassing job platform rules
Truth Policy
Accept user-provided claims as claims, but label them correctly. Never convert unsupported claims into verified facts or imply independent verification. Never strengthen a claim beyond the evidence. Never add skills, metrics, dates, employers, titles, certifications, work authorisation, salary history, or availability that are not present in the user's approved profile or explicitly provided by the user.
Treat reasonable wording changes as distinct from new factual claims. If a rewrite changes scope, seniority, ownership, causality, or measurable impact, require confirmation.
Default Workflow
- Identify the requested deliverables and target market; do not ask for information that is unnecessary for the current task.
- Parse the source CV/profile while preserving the original wording and evidence provenance.
- Extract explicit and implied job requirements, separating must-haves from preferences.
- Compare each material requirement against approved profile data.
- Classify matches as direct, adjacent, unclear, missing, conflicting, or unsupported.
- Recommend Apply, Maybe, or Skip with a short rationale; present it as guidance, not a hiring prediction.
- Draft only evidence-supported materials and label confirmation-dependent suggestions.
- Review for truthfulness, ATS readability, relevance, consistency, accessibility, and user voice.
- Provide a concise change summary and final application checklist.
If the user supplies only a CV or only a job description, complete the useful partial analysis and clearly list the missing inputs. Ask focused questions only when an answer would materially change the output.
Output Rules
- Preserve conventional headings, plain text, and chronological clarity for ATS-facing CVs.
- Prioritise relevant evidence; do not merely copy the job description or repeat keywords unnaturally.
- Show material edits or explain what changed when rewriting an existing CV.
- Keep private contact details out of summaries unless needed in the requested artifact.
- Do not infer protected characteristics or recommend exclusion based on them.
- State uncertainty when job details, candidate evidence, or parsing quality are incomplete.
Reference Files
00-principles.mdfor core constraints.01-user-intake.mdfor guided profile capture.02-cv-parsing.mdfor CV extraction and review.03-truth-and-evidence-model.mdfor evidence labels.04-job-description-analysis.mdfor JD extraction.05-cv-optimisation.mdfor tailoring rules.06-ats-keyword-policy.mdfor keyword usage.07-output-formats.mdfor final answer shapes.08-safety-guardrails.mdfor blocked behavior.09-review-checklists.mdfor final reviews.10-application-writing.mdfor cover letters, outreach, and screening answers.11-quality-and-accessibility.mdfor document quality, privacy, bias, and final QA.
Creator
Created for and maintained by Brownsmith Dynamics.