The Job Search Cheat Code — Scan, Score, Tailor, Apply, Track. All Automated.
The Job Search Cheat Code — Scan, Score, Tailor, Apply, Track. All Automated.Open-source. Free to fork. Your resume rewrites itself per job. Welcome to the job search cheat code.
Someone built an AI that reads job listings, judges if they’re worth your time, rewrites your entire resume to match, fills out the application form, and tracks everything — while you’re eating dinner.
It’s called career-ops. It runs on Claude Code. The guy who built it evaluated 740+ jobs, generated 100+ custom resumes, and landed a Head of Applied AI role. Then he open-sourced the whole thing. The repo has 8,000+ stars and climbing. This isn’t a “paste your resume into ChatGPT” situation — this is a full command center that treats job hunting like an engineering problem.
⚡ Wait — What Does It Actually Do?
Think of it like this: you give it a job listing URL. It does everything else.
| Stage | What Happens | You Do |
|---|---|---|
| Scan | Crawls 45+ company career pages automatically | Nothing |
| Evaluate | Scores every job A-F across 10 dimensions vs YOUR resume | Glance at scores |
| Tailor | Rewrites your resume per job — keywords, structure, framing | Review the PDF |
| Apply | Opens the application form, fills every field, uploads your CV | Click submit |
| Track | Logs everything in one file — status, scores, URLs, dates | Check when curious |
The trick nobody mentions: Career-ops doesn’t just match keywords. It REASONS about whether you’re actually a fit — then rewrites your resume to prove it. Same real experience, 6 different framings for 6 different jobs. Nothing fabricated, everything repositioned.
It has 14 different modes — like cheat codes for different situations:
| Command | What It Does |
|---|---|
/career-ops scan |
Hunts for new openings across all configured companies |
/career-ops pdf |
Generates an ATS-optimized resume as a polished PDF |
/career-ops batch |
Evaluates 10+ jobs in parallel using sub-agents |
/career-ops deep |
Deep-dives a specific company before your interview |
/career-ops apply |
AI fills out the application form for you |
/career-ops contacto |
Drafts personalized LinkedIn outreach messages |
/career-ops training |
Evaluates whether a course/cert is worth your time |
🧠 Use Cases That'll Make You Go 'Wait... You Can Do THAT?'
Forget the obvious “it applies to jobs for you” angle. Here’s what ACTUALLY makes this insane:
1. It builds a STAR story bank from your career — and it gets smarter over time.
Every evaluation generates behavioral interview answers (Situation → Task → Action → Result + Reflection) from YOUR real experience. After 20+ evaluations, you have 5-10 master stories that answer literally any interview question. You never walk into an interview unprepared again.
2. It scores jobs to SAVE you from applying to garbage.
740 evaluations. Maybe 20-30 actual applications. The tool’s real superpower isn’t applying — it’s saying NO to 95% of jobs before you waste emotional energy on them. It’s a mental health tool disguised as a productivity tool.
3. It generates salary negotiation scripts PER offer.
The negotiation mode knows your competing offers, your BATNA, geographic discount pushback language, and exactly when to name a number vs. wait. Pair it with Levels.fyi salary data (they literally have AI-agent-friendly endpoints at /companies/{name}/salaries.md) and you’re negotiating with real comp data, not vibes.
4. The architecture IS the tutorial.
Career-ops uses 14 Claude Code skill files with shared context and isolated modes. If you’re building ANY agentic Claude Code project — customer support bot, content pipeline, research agent — this repo is a masterclass in how to structure multi-mode AI systems. The job search part is almost secondary.
5. You can plug in industrial-grade job scrapers as the discovery layer.
Career-ops scans 45 companies. But python-jobspy scrapes LinkedIn, Indeed, Glassdoor, Google, ZipRecruiter, and Bayt in ONE call. Feed those results into career-ops for evaluation. Suddenly you’re scanning the entire internet, not just 45 career pages.
6. The companion portfolio site is ALSO open-source.
cv-santiago is a forkable portfolio website with an AI chatbot that answers recruiter questions using your resume data. The chain: career-ops finds jobs → tailors your CV → cv-santiago presents you online when they Google your name. Two repos. One complete job search system.
7. Greenhouse, Lever, and Ashby have FREE public APIs that nobody uses.
https://boards-api.greenhouse.io/v1/boards/{company}/jobs returns every job listing as clean JSON. No auth. No scraping. No Playwright. Jobber wraps Greenhouse + Lever + Ashby + BambooHR into one dead-simple API. One fetch call = every open role at any company using these platforms.
🔧 Setup — 5 Minutes, Not 5 Hours
What you need: Node.js, a Claude Code subscription, and 5 minutes.
git clone https://github.com/santifer/career-ops.git
cd career-ops && npm install
npx playwright install chromium
Then open Claude Code in the directory and literally tell it to set itself up:
claude
# "Change the archetypes to backend engineering roles"
# "Translate the modes to English"
# "Add these 5 companies to portals.yml"
# "Update my profile with this CV I'm pasting"
Cost reality check: The creator ran this on Claude Max ($200/mo). You CAN run it on cheaper plans, but batch mode (parallel evaluations) will eat tokens fast. Single evaluations on Pro ($20/mo) work fine. The $200 plan is for “evaluate 50 jobs in one sitting” energy.
🛠️ The Bigger Toolkit — Stack These Together
Career-ops is one piece. Here’s the full open-source job search arsenal:
| Tool | What It Does | Cost | Link |
|---|---|---|---|
| career-ops | Evaluate + tailor CV + apply + track (Claude Code) | Claude sub | 8k+  |
| ApplyPilot | 6-stage autonomous pipeline, solves CAPTCHAs | Free (Gemini API) | Full alternative |
| AIHawk | LinkedIn Easy Apply auto-applier | Free + API key | OG auto-applier |
| python-jobspy | Scrape LinkedIn + Indeed + Glassdoor + 4 more boards | Free | pip install |
| Jobber | Lightweight API for Greenhouse + Lever + Ashby | Free | One fetch call |
| cv-santiago | Portfolio site with AI chatbot, forkable | Free | Companion repo |
The stack that covers everything:
JobSpy discovers → career-ops evaluates and tailors → ApplyPilot or career-ops submits → cv-santiago presents you online.
⚠️ Don't Be Stupid — Read This
| Do | Don’t |
|---|---|
| Review every resume before sending | Blindly auto-submit 200 applications |
| Space out applications — ATS tracks velocity | Batch-fire 20 apps in one session |
| Use it as a FILTER — evaluate 700, apply to 30 | Spray the internet with your resume |
| Rewrite AI-generated interview stories in your voice | Memorize AI answers verbatim |
| Check Levels.fyi before naming a salary | Paste AI negotiation scripts unedited |
What NOT to do: White-font keyword stuffing, prompt injection in resumes, submitting without reviewing. Recruiters detect all of this. Career-ops does keyword REFORMULATION (rewording your real experience to match the JD), not injection (stuffing invisible text). Massive difference.
Quick Hits
| Want | Do |
|---|---|
 Full pipeline |
→ career-ops |
 Free alternative |
→ ApplyPilot — Gemini, zero cost |
 Scrape all job boards |
→ python-jobspy |
 Portfolio + AI chatbot |
→ cv-santiago |
 Real salary data |
→ Levels.fyi — AI-agent endpoints |
 Claude Code ecosystem |
→ awesome-claude-code |
Recruiters automated rejecting you years ago. About time you automated applying.
!