Triggers on "improve my CV", "rewrite resume", "enhance CV", "make my CV better", "CV for tech role", "mejorar mi hoja de vida", "reescribir CV", "CV para entrevista". Produces impact-rewritten CV using PAR framework. Output: markdown + optional branded HTML.
From jm-adknpx claudepluginhub javimontano/jm-adk-alfaThis skill is limited to using the following tools:
agents/guardian.mdagents/lead.mdagents/specialist.mdagents/support.mdevals/evals.jsonknowledge/body-of-knowledge.mdknowledge/knowledge-graph.mdprompts/meta.mdprompts/primary.mdprompts/variations/deep.mdprompts/variations/quick.mdreferences/cv-patterns.mdtemplates/output.docx.mdtemplates/output.htmlSearches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
TL;DR: Audit a CV for structural and language weaknesses, rewrite every bullet using the PAR framework (Problem-Action-Result), optimize for ATS systems, target to a specific role, and deliver a polished markdown CV plus an optional JM Labs dark-branded HTML version.
Principio Rector: Hiring managers read CVs for evidence of impact, not lists of responsibilities. Every bullet must answer: "So what?" If a bullet could belong to anyone in that role at that company, it has not been rewritten yet.
Activate when:
Do NOT activate when:
Before rewriting, diagnose what is weak. Apply a structured rubric to the CV as provided. Share the audit findings with the user before beginning rewrites — confirm scope. [EXPLICIT]
Structure Audit — section presence and quality:
| Section | Required | Quality Check |
|---|---|---|
| Contact info | Yes | Email, LinkedIn, city/country; phone optional; no photo in most markets |
| Professional summary | Recommended | 3–4 lines, role-specific, at least one quantified claim |
| Work experience | Yes | Reverse chronological, impact bullets, dates with months |
| Education | Yes | Degree, institution, graduation year; omit GPA unless >3.7 or recent grad |
| Skills | Yes | Categorized (technical, languages, tools), relevant to target role |
| Projects / portfolio | Role-dependent | Mandatory for tech, design, and product roles |
| Certifications | Optional | Include if relevant and issued within last 5 years |
| Publications / talks | Optional | Include for thought leadership, academic, or senior roles |
Language Quality Audit — patterns to flag and eliminate: [EXPLICIT]
Red flags — language to rewrite:
Green flags — language to preserve or generate:
ATS Readiness Audit: [EXPLICIT]
Audit Output Format (share before rewriting):
CV AUDIT SUMMARY
─────────────────────────────────────────────
Sections present: [list with quality score H/M/L]
Sections missing: [list with recommendation]
Language red flags: [count] found — [2-3 example bullets]
ATS issues: [list of format problems]
Metrics present: [X out of Y bullets have quantified outcomes]
Overall readiness: [score /10]
Recommended scope: [Full rewrite / Targeted edits / Light polish]
─────────────────────────────────────────────
Apply the PAR (Problem-Action-Result) framework to every work experience bullet. [EXPLICIT] This is the core transformation. Do not skip bullets — every line must pass the PAR test.
PAR Structure:
P — Problem / Context: What was the situation, challenge, or goal?
A — Action: What did YOU specifically do? (your contribution, not the team's)
R — Result: What measurably improved? (quantify whenever possible)
Compression rule: Most bullets collapse PAR into one sentence. Full P+A+R structure is used for the most impactful or complex achievements.
PAR Transformation Examples:
| Before (Duty language) | After (PAR-Impact) |
|---|---|
| "Managed social media accounts" | "Grew LinkedIn following from 2K to 18K in 8 months by launching a weekly expert series, driving 340% increase in inbound leads" |
| "Worked on backend API development" | "Engineered REST API handling 50K req/day for fintech product, reducing p95 latency from 800ms to 120ms via Redis caching layer" |
| "Led sales team" | "Led 8-person enterprise sales team to 127% of annual quota ($4.2M ARR), closing 3 Fortune 500 accounts in Q3 2023" |
| "Responsible for data analysis" | "Built automated reporting pipeline (Python + BigQuery) reducing weekly analysis time from 16h to 2h, enabling weekly board reviews" |
| "Helped improve customer satisfaction" | "Redesigned onboarding flow for 12K users, lifting 30-day retention from 41% to 67% and reducing support tickets by 38%" |
Metrics Injection Protocol — when the CV lacks numbers: [INFERRED]
If the user's CV has no metrics, do one of two things:
Scope proxies (no fabrication required):
Targeted questions to prompt user for missing metrics:
[EXPLICIT] Metrics must be truthful. Never fabricate numbers. If the user cannot confirm a metric, use a scope proxy or leave the claim qualitative but specific.
Action Verb Bank by Category:
| Category | Verbs |
|---|---|
| Leadership | Led, Directed, Coached, Mentored, Built, Scaled, Reorganized, Championed, Spearheaded |
| Technical | Engineered, Architected, Developed, Automated, Migrated, Optimized, Deployed, Integrated |
| Analysis | Analyzed, Evaluated, Modeled, Forecasted, Benchmarked, Diagnosed, Quantified, Identified |
| Commercial | Closed, Negotiated, Grew, Expanded, Converted, Retained, Upsold, Acquired, Prospected |
| Operations | Streamlined, Reduced, Eliminated, Standardized, Implemented, Launched, Delivered, Managed |
| Communication | Presented, Facilitated, Published, Authored, Trained, Influenced, Partnered, Translated |
Rewrite each CV section with role-specific and level-appropriate language. [EXPLICIT]
Professional Summary / Profile The summary is the first-read section. Must be: role-specific (not generic), 3–4 lines, and contain at least one quantified claim or scope signal.
Bad: "Experienced professional with strong communication skills and a passion for results." Good: "Senior product manager with 8 years building B2B SaaS products. Led 0-to-1 launch of [Product] reaching $2.4M ARR in 14 months. Specializes in data-driven roadmapping, cross-functional alignment, and enterprise go-to-market. Based in Bogotá, open to remote."
Work Experience Bullets Apply PAR to every bullet. Aim for 3–5 bullets per role. More recent roles get more bullets. Order bullets within a role: most impactful first, not chronological. No bullet should read like a job description. Every bullet should read like evidence of value delivered.
Skills Section Structure as named categories, not a flat undifferentiated list:
Technical: Python, SQL, dbt, Snowflake, Tableau, Looker
Frameworks: Scrum, SAFe, OKRs, Lean Six Sigma
Languages: Spanish (native), English (C1), Portuguese (B2)
Tools: JIRA, Confluence, Salesforce, Figma
Remove entry-level-expected skills for the target role (e.g., "Microsoft Word" for a senior PM). Keep only differentiating or explicitly required skills.
Education
Certifications
Ensure the CV passes automated screening before it reaches a human reviewer. [EXPLICIT]
Keyword Strategy (4-step process):
Keyword Categories by ATS Weight:
| Category | ATS Weight | Examples |
|---|---|---|
| Job title match | Very High | "Product Manager", "Senior Data Engineer" |
| Hard technical skills | High | "Python", "Salesforce", "AWS", "dbt" |
| Certifications | High | "PMP", "AWS Solutions Architect", "CISSP" |
| Industry terms | Medium | "SaaS", "fintech", "HIPAA", "GDPR", "ISO 27001" |
| Methodologies | Medium | "Scrum", "Agile", "Kanban", "Six Sigma" |
| Soft skills | Low | "leadership", "communication" — use sparingly |
Formatting Rules for ATS Compliance:
| Rule | Rationale |
|---|---|
Use .docx or clean PDF | Designer PDFs (Canva, InDesign) often fail ATS parsing |
| No columns or tables for experience | Column parsers split dates from content |
| Spell out acronyms first use | "Machine Learning (ML)" before using "ML" alone |
| Include months AND years | "Jan 2022 – Mar 2024" not "2022 – 2024" |
| Standard date format | Prevents duration calculation errors in ATS |
| No images, icons, or logos in body | ATS ignores or errors on visual elements |
| Standard section headers | "Work Experience" not "Where I've Worked" |
| Left-aligned text | Centered or right-aligned body text can confuse parsers |
Keyword Density Check (after rewriting):
Customize the CV for a specific job posting or industry vertical. [EXPLICIT]
Single-Role Targeting (when user provides a specific posting):
Industry-Specific Calibrations:
| Industry | Key Adjustments |
|---|---|
| Tech / SaaS | Stack-first ordering, quantified engineering impact, GitHub or portfolio link |
| Finance | Regulatory knowledge (SOX, Basel, MiFID), deal size, P&L ownership, CFA/CPA |
| Consulting | Client count, engagement scope, industry verticals served, frameworks used |
| Healthcare | Certifications (HIPAA, HITECH), patient outcomes, EHR systems, compliance |
| Startups | Bias-to-action verbs, generalist range, 0-to-1 initiatives, scrappy resourcefulness |
| Enterprise | Scale, governance, stakeholder management, budget size, cross-org impact |
| Sales | Quota attainment %, deal size, sales cycle length, named accounts, CRM proficiency |
| Product | OKRs owned, revenue impact, user growth metrics, cross-functional leadership |
Career Level Calibrations:
| Level | Summary Focus | Bullet Depth | Format Bias |
|---|---|---|---|
| Junior (0–3y) | Skills, education, projects | Task + tool + learning outcome | 1 page max |
| Mid (3–7y) | Results + growth trajectory | PAR with metrics | 1–2 pages |
| Senior (7–15y) | Leadership + strategy | Business impact + team and org scope | 2 pages |
| Executive (15y+) | Vision + P&L + org transformation | Revenue, company-wide change | 2 pages max |
Career Change Targeting (switching tracks entirely):
Two output formats. Produce markdown by default; produce HTML when user requests it. [EXPLICIT]
Format A: Markdown CV — clean, portable, ATS-safe for digital job submissions:
# [Full Name]
[City, Country] · [email@domain.com] · [linkedin.com/in/handle]
## Professional Summary
[3–4 lines. Role-specific. At least one quantified claim. No first-person pronouns.]
## Experience
### [Job Title] — [Company Name], [City / Remote]
*[Mon YYYY] – [Mon YYYY]*
- [PAR bullet — strongest impact first]
- [PAR bullet — second most relevant]
- [PAR bullet — third]
### [Previous Job Title] — [Company Name], [City]
*[Mon YYYY] – [Mon YYYY]*
- [PAR bullet]
- [PAR bullet]
## Skills
**Technical:** [comma-separated list]
**Frameworks:** [comma-separated list]
**Languages:** [Language (Level), Language (Level)]
**Tools:** [comma-separated list]
## Education
**[Degree, Field]** — [Institution Name], [Year]
## Certifications
- [Certification Name] — [Issuing Body] ([Year])
Format B: Branded HTML CV — JM Labs dark design system, for portfolio sharing.
Self-contained single file. Not for ATS submission. Includes @media print CSS. [EXPLICIT]
Load the full HTML template from references/html-cv-template.md when producing Format B.
JM Labs Brand Tokens (required for HTML output):
:root {
--bg: #0d0d0d;
--card: #1a1a2e;
--navy: #122562;
--gold: #FFD700;
--blue: #137DC5;
--text: #e0e0e0;
--muted: #9e9e9e;
--font-h: 'Poppins', sans-serif;
--font-b: 'Inter', sans-serif;
}
HTML CV Structure (sections in order):
<header> — Name (gold), target role title (blue), contact bar (muted text + linked email/LinkedIn).role card per position (gold title, muted dates, PAR bullet list with gold triangles)<footer> — "Generated with JM Labs CV Enhancement · [DATE]" in muted textBilingual support: Add [data-l="es"] / [data-l="en"] on section titles and labels
for clients who need to switch language of the CV display without re-generating. [EXPLICIT]
Print CSS requirement: Include @media print block that switches to white background,
#122562 for headings, #f8f8f8 for cards. Page-break-inside: avoid on .role blocks. [EXPLICIT]
[EXPLICIT] HTML output is for portfolio sharing and direct recruiter review. For ATS submission, always provide the markdown version as clean PDF or .docx. HTML-only CVs will not parse correctly in most ATS systems.
| Decision | Option A | Option B | Recommendation |
|---|---|---|---|
| Impact language vs. factual accuracy | Maximize impact phrasing | Stay strictly literal | Impact is fine; fabricating numbers is not. Scope proxies resolve the gap. |
| ATS optimization vs. human readability | Dense keywords, standard format | Natural writing, design-forward | ATS-first for applications; human-first for networking/sharing versions |
| Breadth vs. depth per role | Many bullets, wide coverage | 3–5 bullets, highest impact only | Fewer, stronger bullets always outperform long lists of mediocre ones |
| Generalist vs. targeted summary | Generic summary for all roles | Role-specific summary per application | Role-specific always wins for conversion rate |
| One-page vs. two-page | Strict one page | Two pages for senior | Junior: one page max. Mid and above: two pages fine. Executive: two pages max. |
| Markdown vs. HTML output | Clean markdown (ATS-safe) | Branded HTML (visual, shareable) | Markdown for job applications; HTML for portfolio, LinkedIn share, recruiter outreach |
| Scenario | Handling |
|---|---|
| CV in Spanish, target role in English | Translate + adapt: idioms, title conventions, and metric framing differ by market. Produce English version; offer bilingual HTML if requested. |
| Career gap of >12 months | Address directly in the summary with positive framing: freelance, caregiving, further education, sabbatical. Do not hide or obscure the gap. |
| Career change (different track entirely) | Switch to skills-first format. Add "Relevant Projects" section. Reframe past experience in the target industry's language. Summary leads with the pivot narrative. |
| Academic CV (publications-heavy) | Switch to academic format conventions. Do not strip the publications section. Add teaching experience, grants, and committee roles if present. |
| New graduate / <1 year experience | Amplify projects, coursework, internships, hackathons, volunteer work. Education section leads over experience. One page strictly. |
| Long tenure in one role (8y+ same company) | Break the role into sub-roles or initiative clusters if scope changed significantly. Prevents appearance of stagnation. |
| User refuses to provide any metrics | Use scope proxies throughout. Note in output that metrics are scope-based, not outcome-based. |
BAD — Duty language, no impact:
"Responsible for managing the development team and ensuring project delivery."
GOOD — PAR-transformed, specific, quantified:
"Led 6-person cross-functional engineering team delivering 3 major product releases in 2023, reducing deployment cycle from 3 weeks to 5 days through CI/CD pipeline implementation (GitHub Actions + Docker)."
Analysis of the transformation:
Before delivering any CV output, verify all of the following:
#122562, gold #FFD700, blue #137DC5, dark bg)@media print CSS and [data-l] bilingual attributes on section labels| File | Purpose |
|---|---|
knowledge/body-of-knowledge.md | CV writing methodology and PAR framework research |
references/html-cv-template.md | Full JM Labs branded HTML CV template |
evals/ | Before/after CV examples with scoring rubric |
landing-pages — Personal brand page to complement the CVb2b-outreach — Craft outreach messages using CV highlights as professional credentialsexecutive-pitch — Translate career story into a compelling 5-minute verbal pitchux-writing — Microcopy principles applicable to tight, scannable bullet writing