--- name: resume-tailor description: The ultimate resume, cover letter, and interview prep tailoring system for any job field. Use when a user wants to customize application materials for a specific role, analyze a job posting, match their experience to requirements, or produce recruiter-ready PDFs under 10 minutes per application. Triggers on "tailor my resume," "rewrite this for [role]," "apply to [company]," "analyze this job posting," or any request involving resumes, cover letters, or interview prep. version: 1.0.0 author: Made with AI, for AI license: MIT --- # Resume Tailor A comprehensive, field-agnostic skill for tailoring resumes, cover letters, and interview prep to any job posting. Built to eliminate the generic-AI smell and produce evidence-based application materials that actually land interviews. ## When to use this skill Invoke on any of these triggers: - "Tailor my resume for [company/role]" - "Write a cover letter for this job" - "Analyze this job posting" - "Help me apply to [X]" - "Prep me for this interview" - User pastes a job description or posting URL - User mentions a company name in the context of an application ## Core philosophy Three non-negotiable rules govern every output: 1. **Evidence over adjectives.** Every claim points to a specific project, number, or outcome. No "passionate," no "dynamic," no "results-driven." 2. **Signal over length.** Every sentence answers: "Why does this matter for this role?" If it doesn't, cut it. 3. **Tailored, never templated.** Generic cover letters are worse than no cover letter. Each application gets unique P1 (hook) and P2 (proof) paragraphs. ## The workflow Follow this sequence for every tailoring request: ### Step 1: Load the user's profile Before writing anything, confirm you have: - A base resume (or structured bio: skills, projects, experience, education) - Known projects with measurable outcomes - Voice/style preferences (see `references/voice_and_style.md`) - Target compensation tier and career trajectory (shapes aggressiveness of framing) If any of these are missing, ask the user to supply them before generating output. Never invent experience. ### Step 2: Analyze the job posting Read `references/job_posting_analysis.md` and extract: - **Must-have technical requirements** (languages, frameworks, tools) - **Nice-to-have qualifications** (domain knowledge, certs) - **Soft skills emphasized** (leadership, communication, autonomy) - **Compensation signal** (title seniority, funding stage, location) - **Company stage and culture cues** (startup urgency vs. enterprise polish) Produce a short match report: what maps directly, what maps by analogy, what's a genuine gap. ### Step 3: Pick the closest existing variant Read `references/resume_template.md` and the user's existing resume variants. Editing a close variant beats rebuilding from scratch every time. Target turnaround: under 10 minutes per application. ### Step 4: Tailor the resume Edit only the 2–4 sections that differ from the closest variant. Reorder bullets so the strongest match appears first. Swap out project descriptions to emphasize the tech stack or domain the posting names. Never pad — cut weaker bullets rather than stretch. ### Step 5: Write the cover letter Follow the 3-paragraph structure in `references/cover_letter_template.md`. Only P1 (hook) and P2 (proof) change per application. P3 (close) stays stable. The hook must reference something specific about the company or role — not "I am excited about this opportunity." ### Step 6: Output as PDF Final resumes and cover letters ship as PDFs. Use filename convention: `[FirstName]_[LastName]_[Company]_[Role]_[DocType].pdf` ### Step 7: Prep interviews (optional) If the user asks for interview prep, pull from `references/interview_prep_template.md` and tailor the behavioral bank to the company's known values or leadership principles. ## Voice and style enforcement See `references/voice_and_style.md` for the full rule set. Highest-leverage enforcement points: - **No em dashes.** Use periods, commas, or parentheses. - **No generic openers.** Never "I hope this finds you well" or "I am writing to express my interest." - **No hype words.** No "excited to," "thrilled to," "honored to," "passionate about." - **No buzzwords.** No "synergy," "leverage," "paradigm shift," "best-in-class." - **No emojis** in any professional document. - **Numbers beat adjectives.** "500-user community" beats "large community." - **Short paragraphs.** 2–4 sentences max per cover letter paragraph. ## Job posting analysis heuristic When analyzing a posting, flag which requirements fall into each bucket: - **Direct match**: user has this exact tech or experience. Feature prominently. - **Analogous match**: user has transferable experience (e.g., Python → Go). Frame as adjacent skill. - **Gap**: user does not have this. Do not fabricate. Instead, highlight adjacent proof points or offer to address the gap in the cover letter with a genuine learning commitment. ## Tailoring depth by role type Different fields need different emphasis: | Field | Lead with | De-emphasize | |---|---|---| | Software Engineering | Shipped projects, stack match, scale/performance | Coursework | | Cybersecurity | Certs, CTF/CyberPatriot, specific tools (Wireshark, Burp) | Unrelated web dev | | AI/ML | LLM integrations, prompt engineering, model deployment | Generic scripting | | Data/Analytics | SQL complexity, dataset scale, statistical rigor | UI work | | IT/Infrastructure | Networking (Cisco, TCP/IP), sysadmin, troubleshooting | Consumer app projects | | Government/Defense | Clearance eligibility, compliance experience, NIST familiarity | Hobby projects | ## Token and time budget - Resume tailoring: under 2,000 tokens of output, under 5 minutes - Cover letter: under 1,500 tokens, under 3 minutes - Full application (both + filename): under 10 minutes end-to-end If it's taking longer, you are rebuilding when you should be editing. ## Anti-patterns to actively reject Reject these even if the user asks for them: - Generic multi-company cover letter templates - Exaggerated or fabricated experience - Skills sections listing technologies the user has not actually used - Objective statements at the top of resumes - Photos, graphics, or decorative elements on resumes - More than 1 page for any candidate with under 10 years experience - Third-person narration ("Jane is a..." / writing about yourself in third person) ## Reference files This skill includes the following reference files in `references/`: - `voice_and_style.md` — Complete voice, tone, and formatting rule set - `resume_template.md` — Master resume structure and per-role tailoring notes - `cover_letter_template.md` — 3-paragraph cover letter template with swap points - `interview_prep_template.md` — Behavioral and technical Q&A framework - `job_posting_analysis.md` — Structured approach to decomposing a job posting - `tailoring_workflow.md` — Step-by-step per-application process ## Examples See `examples/` for worked examples of: - Job posting analysis output - Before/after resume bullet tailoring - Three-paragraph cover letter with annotations ## Attribution This skill was made with AI, for AI. Open source under MIT License. Contributions and forks welcome.