Stop Rewriting Resumes.
Build a System That Does It For You.

The Problem with Resume Writing (and Why AI Usually Fails At It)

Most AI resume tools are glorified text reformatters. You paste in a job description, they shuffle your bullet points around, and you get something that looks tailored but is really just rearranged vagueness.

The root problem isn't formatting — it's evidence. You can't write a compelling, specific resume bullet six months after a project ends. The metrics fade. The decisions blur. The "proudest achievement" becomes "led cross-functional team to deliver platform migration."

"The best resume writers don't have better resumes. They have better notes."

This guide shows you how to build a system that solves the root problem: capture experience immediately after each project in rich, structured detail — then query that growing library whenever you need a resume, with AI doing the evidence-mapping and tailoring automatically.

System Architecture at a Glance

Two skills, one growing library, unlimited resumes.

Project Structure

Clone the repo, and you get a ready-made folder convention that keeps skills, experience, and resumes cleanly separated:

Cowork-ResumeBuilder/
│
├── Skills/
│   ├── experience-letter-v2/        ← Skill 1: capture experience
│   │   ├── SKILL.md                  ← full interview + generation instructions
│   │   └── references/               ← YAML schema template + filled example
│   │
│   └── resume-enhancer-v2/         ← Skill 2: generate tailored resumes
│       ├── SKILL.md                  ← 8-step resume pipeline
│       └── scripts/
│           └── generate_resume.js    ← .docx template engine (Node.js)
│
├── Resumes/                         ← Base resumes by role level
│   ├── CISO & Executive/
│   ├── Director/
│   ├── Engineering Manager/
│   ├── Principal & Staff/
│   └── Technical IC/
│
├── Experience/                      ← Your growing library (one folder per tenure)
│   └── Security Engineer (2021–Current)/
│       ├── project-alpha.yaml
│       └── Experience_Letter_Alpha.docx
│
└── career_timeline.json            ← canonical index: roles, skills, certs

Skill 1 — Experience Letter Capture

What Makes the YAML Special: Provenance Anchors

Every claimable item in the YAML gets an immutable ID — an anchor. This means every resume bullet can be traced back to the exact evidence that supports it. No hallucination. No inflated claims. Full auditability.

meta:
  project_name: "CSPM Transformation"
  role: "Engineering Manager"
  duration: "Jan 2023 – Dec 2023"

impact:
  quantitative:
    - id: &cspm-attack-surface   # ← provenance anchor
      metric: "Attack surface reduced"
      value: 68
      unit: "%"
      confidence: "measured"
    - id: &cspm-alert-reduction
      metric: "Alert noise reduction"
      value: 40
      unit: "%"
      confidence: "estimate"

Skill 2 — Resume Enhancer

Why This Is Different from "ChatGPT, Write My Resume"

4 Design Insights Worth Stealing

How to Build It in 4 Steps

1. Populate career_timeline.json — Fill in your roles, dates, skills taxonomy, education, and certifications. This is the index all downstream skills query. Start minimal: name, employer, tenure dates, and a skills list per role.

2. Run Skill 1 on 2–3 projects — Build your experience library before touching the Resume Enhancer. Say "Capture my experience on [project name]" and answer Claude's questions honestly. Rough numbers beat no numbers.

3. Add base resumes to Resumes/ — One per role level you're targeting. Even a rough draft Word doc is fine — the enhancer uses it as a formatting baseline.

4. Run Skill 2 when applying — Paste the job description and say "Build my resume for [role]. Here's the JD. Use my Experience folder." Review the evidence map Claude confirms before it writes bullets. Approve edits from the persona review.

The career_timeline.json — Your Career's Source of Truth

This single JSON file is the canonical index that both skills query. It holds your full professional history: roles, tenures, engagement IDs, skill taxonomies, education, and certifications.

{
  "candidate": {
    "name": "Your Name",
    "email": "you@example.com",
    "linkedin": "linkedin.com/in/you"
  },
  "tenures": [
    {
      "employer": "Acme Corp",
      "role_title": "Engineering Manager",
      "start": "2021-01",
      "end": "present",
      "engagements": [
        {
          "id": "cspm-transformation",
          "name": "CSPM Transformation",
          "yaml_path": "Experience/Security Engineer/cspm-transformation.yaml",
          "skills": {
            "tools":    ["Wiz", "Terraform", "AWS Security Hub"],
            "concepts": ["Cloud Security", "CSPM", "Risk Reduction"],
            "categories": ["Cloud Security", "Platform Engineering"]
          }
        }
      ]
    }
  ],
  "certifications": ["CISSP", "AWS Security Specialty"]
}

What to Build Next

The same pattern — focused skill + structured data = compounding value — works for three obvious extensions:

The Real Value: Compounding

The first time you run this system, you'll get a better resume. The tenth time, you'll have a library of richly documented projects that makes every future application trivially easy.

Most people treat resume writing as a one-off task. The professionals who advance fastest treat it as a system — and update it continuously. This is that system.

"Capture experience once, richly, right after a project ends → query that library whenever you apply. The more you capture, the better every future resume gets."