Resume Tailor Skill
Orchestrates a sequential pipeline of specialized agents to analyze a resume against a job description, then produces a tailored resume with gap analysis, ATS scoring, change tracking, and actionable next steps.
Commands
/resume tailor — Full pipeline: analysis, gap report, rewrite, diff, ATS score, synthesis
/resume gap — Gap analysis only (no rewrite)
/resume ats — ATS compatibility score only
/resume cover — Generate cover letter (3 variants: standard, short, talking points)
/resume interview — Interview prep from gap analysis + JD
/resume compare — Compare resume against multiple JDs with fit matrix
Input Requirements
Before starting, collect:
- Resume — paste as text, or provide a
.pdf/.docx file path
- Job Description — paste full JD text or provide a file path
- Target role title (if not clear from JD)
- Optional: Company name, submission channel (ATS portal / email / referral), specific concerns ("I lack X skill")
If any required inputs are missing, ask for them before proceeding. Do NOT proceed with partial inputs.
For /resume compare, collect 2-5 job descriptions instead of one.
Full Tailor Flow (/resume tailor)
Step 1: Structured Analysis (Replaces Python Parser)
Analyze the resume and JD directly in-context. Do NOT use any external scripts. Extract the following structured data by reading both documents carefully:
From the Resume:
- All technical skills (languages, frameworks, tools, platforms, databases, methodologies) — extract everything mentioned, not just from a predefined list
- All soft skills and leadership indicators (team size, mentoring, cross-functional work)
- Experience timeline: for each role, extract company, title, start date, end date. Calculate cumulative experience handling overlapping roles correctly (overlapping months count once, not twice)
- Section-by-section breakdown: summary, experience, skills, education, projects, certifications
- Quantified achievements: any numbers, percentages, scale indicators
From the Job Description:
- All requirements classified as:
- Core Technical — Primary tech stack (weighted highest)
- Supporting Technical — Infrastructure, tooling, adjacent tech
- Process & Methodology — Agile, code review, testing practices
- Soft Skills — Communication, leadership, collaboration
- Nice-to-Have — Explicitly optional items
- Boilerplate — Generic filler ("team player", "fast-paced environment") — ignore these
- Years of experience required
- Seniority signals (IC vs lead vs manager)
- Domain/industry requirements
Keyword Overlap (with semantic matching):
- Exact matches between resume and JD
- Terminology mismatches (same skill, different name — e.g., "JS" vs "JavaScript")
- Truly missing skills (not in resume at all)
- Semantic matches (e.g., resume says "led a team of 5" → JD says "leadership experience")
Store this analysis in memory for use by all downstream agents.
Step 2: Gap Analyzer Agent
Read agents/gap-analyzer.md and run it with:
- Full resume text
- Full JD text
- The structured analysis from Step 1
This agent produces:
- Candidate profile (career stage, move type)
- JD requirement classification (what's truly required vs aspirational)
- Hard gaps with severity ratings (DEAL-BREAKER / SIGNIFICANT / MINOR)
- Soft gaps with rewrite suggestions
- Keyword gaps (true gaps vs terminology mismatches)
- Strengths to emphasize
- Overall match assessment
The gap analysis output feeds all downstream agents. This step must complete before proceeding.
Step 3: Bullet Rewriter Agent
Read agents/bullet-rewriter.md and run it with:
- Full resume text
- Full JD text
- Gap analysis output from Step 2
This agent produces:
- Rewritten summary (if applicable)
- Rewritten experience bullets using XYZ formula
- Rewritten skills section with JD-matched terminology
- List of keywords injected and where
- Warnings about estimated numbers needing verification
Step 4: Diff Generator Agent
Read agents/diff-generator.md and run it with:
- Original resume text
- Tailored resume text (constructed from bullet rewriter output)
- Gap analysis output
- Bullet rewriter output
This agent produces:
- Section-by-section diff of all changes
- Change type classification (rewrite, reorder, add, remove, terminology)
- Reason for each change tied to JD requirements
- Confidence ratings and verification flags
- Summary statistics (X bullets rewritten, Y keywords added, etc.)
Step 5: ATS Scorer Agent
Read agents/ats-scorer.md and run it with:
- Original resume text
- Tailored resume text
- Full JD text
This agent produces:
- Before/after ATS scores with dimension breakdown
- Complete keyword audit table
- Formatting issues
- File format recommendations
- Specific improvement recommendations
Step 6: Synthesize & Output
After all agents complete, merge their outputs into the final deliverable following the Output Format below. Apply these validation rules during synthesis:
Validation Rules:
- No fabrication — Cross-check every rewritten bullet against the original. If a rewrite adds experience, skills, or achievements not present or inferable from the original, reject it and use a conservative rewrite.
- Conflict resolution — If the gap analyzer says something is a hard gap but the bullet rewriter added it, flag the conflict. Hard gaps cannot be resolved by rewriting — they must be addressed in the cover letter or action items.
- Confidence annotations — Any change with
low confidence or [verify] tags must appear in the verification checklist.
- Score consistency — The ATS score improvement should be consistent with the number and type of changes made. A 5-point improvement with 15 keyword additions doesn't add up — investigate.
Gap Only Flow (/resume gap)
- Structured analysis (Step 1)
- Gap Analyzer agent (Step 2)
- Output: Gap report section only (from Output Format)
ATS Only Flow (/resume ats)
- Structured analysis (Step 1)
- ATS Scorer agent (Step 5, using original resume only — no "tailored" score)
- Output: ATS dashboard section only (from Output Format)
Cover Letter Flow (/resume cover)
- If no prior
/resume tailor run in this session, run Steps 1-2 first (need gap analysis)
- Read
agents/cover-writer.md and run it with:
- Tailored resume (or original if no tailor run)
- Full JD text
- Gap analysis output
- Company name and role title
- Output: All three cover letter variants (standard, short, talking points)
Interview Prep Flow (/resume interview)
- If no prior
/resume tailor run in this session, run Steps 1-2 first (need gap analysis)
- Read
agents/interview-prep.md and run it with:
- Gap analysis output
- Full JD text
- Tailored resume (or original)
- Role title and company name
- Output: Technical questions, gap defense scripts, behavioral questions, questions to ask, red flag preparation
Multi-JD Comparison Flow (/resume compare)
- Collect 2-5 job descriptions
- Structured analysis (Step 1) for each JD
- Read
agents/jd-comparator.md and run it with:
- Full resume text
- All JD texts with role titles and company names
- Output: Fit matrix, rankings, shared requirements, skill investment recommendations, application strategy
Output Format
The full /resume tailor output must include ALL of the following sections in this order:
1. Executive Summary
One paragraph containing:
- Overall fit percentage
- Top strength (the single best match between resume and JD)
- Critical gap (the most important thing missing, or "none" if strong match)
- Verdict: one of
STRONG_MATCH, STRONG_WITH_GAPS, MODERATE_MATCH, WEAK_MATCH, POOR_FIT
- One-sentence recommendation (apply as-is, tailor and apply, apply with strong cover letter, consider upskilling first)
2. ATS Score Dashboard
ATS COMPATIBILITY
=================
Original Tailored Change
Keyword Match ██████░░░░ ████████░░ 52% → 85% (+33)
Formatting ████████░░ ████████░░ 80% → 80% (+0)
Completeness ███████░░░ ████████░░ 70% → 80% (+10)
Title Alignment ████░░░░░░ █████████░ 40% → 90% (+50)
─────────────────────────────────────────────────────────
TOTAL ██████░░░░ ████████░░ 58% → 82% (+24)
Include the dimension breakdown explanation from the ATS scorer.
3. Gap Report
Ordered by severity:
DEAL-BREAKER gaps (if any):
These gaps may prevent your application from advancing. Address in cover letter and consider upskilling.
SIGNIFICANT gaps:
Important but addressable. Covered by resume tailoring and cover letter.
MINOR gaps:
Low risk. Nice-to-haves or items where your adjacent experience is sufficient.
For each gap: requirement, severity, current status, mitigation strategy.
4. Tailored Resume
The complete tailored resume text, ready to copy-paste. Clean formatting, no JSON, no metadata — just the resume as it should be submitted.
5. Change Log
A table showing every modification:
| # | Section | Change Type | Original | New | Reason | Confidence |
|---|---------|-------------|----------|-----|--------|------------|
| 1 | Summary | add | (none) | "Senior Backend..." | Match JD title | high |
| 2 | Exp - Fooddarzee #1 | rewrite | "Built REST APIs" | "Designed RESTful APIs..." | Add specificity + keywords | medium |
Flag items needing verification with a marker.
6. Verification Checklist
Numbered list of every item the candidate must verify before submitting:
- Estimated numbers that need confirmation
- Reframes that might overstate experience
- Terminology changes that need validation
7. Action Items
Numbered checklist of next steps:
[ ] Review and verify all [verify] tagged items in the tailored resume
[ ] Update resume file with tailored version
[ ] Submit via [recommended format] to [ATS/email]
[ ] Prepare cover letter addressing [critical gap]
[ ] Consider: [skill investment recommendation]
[ ] Run /resume interview for interview preparation
Principles
- Never fabricate experience or skills — reframe what exists, flag what doesn't
- Preserve the candidate's voice — don't over-sanitize into generic corporate speak
- Flag hard gaps honestly — if a requirement genuinely can't be addressed, say so
- Prioritize recency — most relevant/recent experience gets the most attention
- Every change needs a reason — no modifications without a clear JD-driven rationale
- Candidate has final say — the verification checklist exists because the candidate must review and approve every change
1---2name: resume-tailor3description: Tailor a resume to a specific job description. Spawns sequential agents to analyze gaps, rewrite bullets, generate diffs, and score ATS compatibility. Trigger when user says "tailor my resume", "match my resume to this JD", "help me apply for this job", "update my resume for this role", or provides a job description alongside their resume.4metadata:5 source: https://skillmd.com/skills/nishilbhave/resume-tailor6---78# Resume Tailor Skill910Orchestrates a sequential pipeline of specialized agents to analyze a resume against a job description, then produces a tailored resume with gap analysis, ATS scoring, change tracking, and actionable next steps.1112## Commands1314- `/resume tailor` — Full pipeline: analysis, gap report, rewrite, diff, ATS score, synthesis15- `/resume gap` — Gap analysis only (no rewrite)16- `/resume ats` — ATS compatibility score only17- `/resume cover` — Generate cover letter (3 variants: standard, short, talking points)18- `/resume interview` — Interview prep from gap analysis + JD19- `/resume compare` — Compare resume against multiple JDs with fit matrix2021---2223## Input Requirements2425Before starting, collect:261. **Resume** — paste as text, or provide a `.pdf`/`.docx` file path272. **Job Description** — paste full JD text or provide a file path283. **Target role title** (if not clear from JD)294. **Optional:** Company name, submission channel (ATS portal / email / referral), specific concerns ("I lack X skill")3031If any required inputs are missing, ask for them before proceeding. Do NOT proceed with partial inputs.3233For `/resume compare`, collect 2-5 job descriptions instead of one.3435---3637## Full Tailor Flow (`/resume tailor`)3839### Step 1: Structured Analysis (Replaces Python Parser)4041Analyze the resume and JD directly in-context. Do NOT use any external scripts. Extract the following structured data by reading both documents carefully:4243**From the Resume:**44- All technical skills (languages, frameworks, tools, platforms, databases, methodologies) — extract everything mentioned, not just from a predefined list45- All soft skills and leadership indicators (team size, mentoring, cross-functional work)46- Experience timeline: for each role, extract company, title, start date, end date. Calculate cumulative experience handling overlapping roles correctly (overlapping months count once, not twice)47- Section-by-section breakdown: summary, experience, skills, education, projects, certifications48- Quantified achievements: any numbers, percentages, scale indicators4950**From the Job Description:**51- All requirements classified as:52 - **Core Technical** — Primary tech stack (weighted highest)53 - **Supporting Technical** — Infrastructure, tooling, adjacent tech54 - **Process & Methodology** — Agile, code review, testing practices55 - **Soft Skills** — Communication, leadership, collaboration56 - **Nice-to-Have** — Explicitly optional items57 - **Boilerplate** — Generic filler ("team player", "fast-paced environment") — ignore these58- Years of experience required59- Seniority signals (IC vs lead vs manager)60- Domain/industry requirements6162**Keyword Overlap (with semantic matching):**63- Exact matches between resume and JD64- Terminology mismatches (same skill, different name — e.g., "JS" vs "JavaScript")65- Truly missing skills (not in resume at all)66- Semantic matches (e.g., resume says "led a team of 5" → JD says "leadership experience")6768Store this analysis in memory for use by all downstream agents.6970### Step 2: Gap Analyzer Agent7172Read `agents/gap-analyzer.md` and run it with:73- Full resume text74- Full JD text75- The structured analysis from Step 17677This agent produces:78- Candidate profile (career stage, move type)79- JD requirement classification (what's truly required vs aspirational)80- Hard gaps with severity ratings (DEAL-BREAKER / SIGNIFICANT / MINOR)81- Soft gaps with rewrite suggestions82- Keyword gaps (true gaps vs terminology mismatches)83- Strengths to emphasize84- Overall match assessment8586**The gap analysis output feeds all downstream agents. This step must complete before proceeding.**8788### Step 3: Bullet Rewriter Agent8990Read `agents/bullet-rewriter.md` and run it with:91- Full resume text92- Full JD text93- Gap analysis output from Step 29495This agent produces:96- Rewritten summary (if applicable)97- Rewritten experience bullets using XYZ formula98- Rewritten skills section with JD-matched terminology99- List of keywords injected and where100- Warnings about estimated numbers needing verification101102### Step 4: Diff Generator Agent103104Read `agents/diff-generator.md` and run it with:105- Original resume text106- Tailored resume text (constructed from bullet rewriter output)107- Gap analysis output108- Bullet rewriter output109110This agent produces:111- Section-by-section diff of all changes112- Change type classification (rewrite, reorder, add, remove, terminology)113- Reason for each change tied to JD requirements114- Confidence ratings and verification flags115- Summary statistics (X bullets rewritten, Y keywords added, etc.)116117### Step 5: ATS Scorer Agent118119Read `agents/ats-scorer.md` and run it with:120- Original resume text121- Tailored resume text122- Full JD text123124This agent produces:125- Before/after ATS scores with dimension breakdown126- Complete keyword audit table127- Formatting issues128- File format recommendations129- Specific improvement recommendations130131### Step 6: Synthesize & Output132133After all agents complete, merge their outputs into the final deliverable following the Output Format below. Apply these validation rules during synthesis:134135**Validation Rules:**1361. **No fabrication** — Cross-check every rewritten bullet against the original. If a rewrite adds experience, skills, or achievements not present or inferable from the original, reject it and use a conservative rewrite.1372. **Conflict resolution** — If the gap analyzer says something is a hard gap but the bullet rewriter added it, flag the conflict. Hard gaps cannot be resolved by rewriting — they must be addressed in the cover letter or action items.1383. **Confidence annotations** — Any change with `low` confidence or `[verify]` tags must appear in the verification checklist.1394. **Score consistency** — The ATS score improvement should be consistent with the number and type of changes made. A 5-point improvement with 15 keyword additions doesn't add up — investigate.140141---142143## Gap Only Flow (`/resume gap`)1441451. Structured analysis (Step 1)1462. Gap Analyzer agent (Step 2)1473. Output: Gap report section only (from Output Format)148149---150151## ATS Only Flow (`/resume ats`)1521531. Structured analysis (Step 1)1542. ATS Scorer agent (Step 5, using original resume only — no "tailored" score)1553. Output: ATS dashboard section only (from Output Format)156157---158159## Cover Letter Flow (`/resume cover`)1601611. If no prior `/resume tailor` run in this session, run Steps 1-2 first (need gap analysis)1622. Read `agents/cover-writer.md` and run it with:163 - Tailored resume (or original if no tailor run)164 - Full JD text165 - Gap analysis output166 - Company name and role title1673. Output: All three cover letter variants (standard, short, talking points)168169---170171## Interview Prep Flow (`/resume interview`)1721731. If no prior `/resume tailor` run in this session, run Steps 1-2 first (need gap analysis)1742. Read `agents/interview-prep.md` and run it with:175 - Gap analysis output176 - Full JD text177 - Tailored resume (or original)178 - Role title and company name1793. Output: Technical questions, gap defense scripts, behavioral questions, questions to ask, red flag preparation180181---182183## Multi-JD Comparison Flow (`/resume compare`)1841851. Collect 2-5 job descriptions1862. Structured analysis (Step 1) for each JD1873. Read `agents/jd-comparator.md` and run it with:188 - Full resume text189 - All JD texts with role titles and company names1904. Output: Fit matrix, rankings, shared requirements, skill investment recommendations, application strategy191192---193194## Output Format195196The full `/resume tailor` output must include ALL of the following sections in this order:197198### 1. Executive Summary199200One paragraph containing:201- Overall fit percentage202- Top strength (the single best match between resume and JD)203- Critical gap (the most important thing missing, or "none" if strong match)204- Verdict: one of `STRONG_MATCH`, `STRONG_WITH_GAPS`, `MODERATE_MATCH`, `WEAK_MATCH`, `POOR_FIT`205- One-sentence recommendation (apply as-is, tailor and apply, apply with strong cover letter, consider upskilling first)206207### 2. ATS Score Dashboard208209```210ATS COMPATIBILITY211=================212 Original Tailored Change213Keyword Match ██████░░░░ ████████░░ 52% → 85% (+33)214Formatting ████████░░ ████████░░ 80% → 80% (+0)215Completeness ███████░░░ ████████░░ 70% → 80% (+10)216Title Alignment ████░░░░░░ █████████░ 40% → 90% (+50)217─────────────────────────────────────────────────────────218TOTAL ██████░░░░ ████████░░ 58% → 82% (+24)219```220221Include the dimension breakdown explanation from the ATS scorer.222223### 3. Gap Report224225Ordered by severity:226227**DEAL-BREAKER gaps** (if any):228> These gaps may prevent your application from advancing. Address in cover letter and consider upskilling.229230**SIGNIFICANT gaps:**231> Important but addressable. Covered by resume tailoring and cover letter.232233**MINOR gaps:**234> Low risk. Nice-to-haves or items where your adjacent experience is sufficient.235236For each gap: requirement, severity, current status, mitigation strategy.237238### 4. Tailored Resume239240The complete tailored resume text, ready to copy-paste. Clean formatting, no JSON, no metadata — just the resume as it should be submitted.241242### 5. Change Log243244A table showing every modification:245246```247| # | Section | Change Type | Original | New | Reason | Confidence |248|---|---------|-------------|----------|-----|--------|------------|249| 1 | Summary | add | (none) | "Senior Backend..." | Match JD title | high |250| 2 | Exp - Fooddarzee #1 | rewrite | "Built REST APIs" | "Designed RESTful APIs..." | Add specificity + keywords | medium |251```252253Flag items needing verification with a marker.254255### 6. Verification Checklist256257Numbered list of every item the candidate must verify before submitting:258- Estimated numbers that need confirmation259- Reframes that might overstate experience260- Terminology changes that need validation261262### 7. Action Items263264Numbered checklist of next steps:265266```267[ ] Review and verify all [verify] tagged items in the tailored resume268[ ] Update resume file with tailored version269[ ] Submit via [recommended format] to [ATS/email]270[ ] Prepare cover letter addressing [critical gap]271[ ] Consider: [skill investment recommendation]272[ ] Run /resume interview for interview preparation273```274275---276277## Principles278279- **Never fabricate experience or skills** — reframe what exists, flag what doesn't280- **Preserve the candidate's voice** — don't over-sanitize into generic corporate speak281- **Flag hard gaps honestly** — if a requirement genuinely can't be addressed, say so282- **Prioritize recency** — most relevant/recent experience gets the most attention283- **Every change needs a reason** — no modifications without a clear JD-driven rationale284- **Candidate has final say** — the verification checklist exists because the candidate must review and approve every change