Santiago Fernández de Valderrama was job hunting. Instead of opening another spreadsheet, he turned Claude Code into a full-blown job search command center. The result: Career-Ops, now sitting at 8.1K GitHub stars and climbing.
He used it to evaluate 740+ job offers, generate 100+ tailored resumes, and land a Head of Applied AI role. The tool worked so well he open-sourced it.
What It Actually Does
Career-Ops runs as a Claude Code skill system with 14 modes. The core loop: you feed it a job description, it scores the fit across 10 weighted dimensions (salary, tech stack, growth potential, etc.) on an A-F scale, then generates an ATS-optimized resume specifically tuned for that listing — not just keyword-stuffed, but with experience reordered by relevance.
The clever part is Playwright integration. Career-Ops autonomously browses Greenhouse, Ashby, Lever, and company career pages, scraping job listings without you copying and pasting URLs. Batch mode lets it process 10+ job descriptions in parallel using sub-agents.
Human stays in the loop. AI analyzes, you decide. It’s a filter that surfaces the 5% of offers worth your time.
How It Compares to LazyApply and Sonara
LazyApply ($99-249/month) auto-submits 50-300 applications per day. Sonara ($5.95/month) does the same on autopilot. Both optimize for volume — spray and pray.
Career-Ops takes the opposite approach. It’s free, open-source, and optimizes for precision. Instead of applying to everything, it tells you which jobs are actually worth applying to before you spend a minute on them. The A-F scoring system means you’re not wasting energy on C-tier matches.
The trade-off: Career-Ops requires Claude Code (meaning an Anthropic API key and some terminal comfort), while LazyApply and Sonara are point-and-click SaaS. If you’re a developer, Career-Ops is clearly the better tool. If you’re not, it’s not for you.
Why Developers Care
Career-Ops is one of the best examples of Claude Code’s skill system being used outside of coding. 14 custom skills, sub-agent orchestration, browser automation, PDF generation — all wired together through CLAUDE.md configuration files. It’s a reference architecture for anyone building multi-skill Claude Code agents.
The fact that the author actually used it to get hired is the strongest product demo possible. No synthetic benchmarks, no cherry-picked screenshots — just 740 evaluated offers and a job offer.
You Might Also Like
- 27k Github Stars in Weeks Learn Claude Code by Shareai lab Breaks Down ai Coding Agents Into 12 Lessons
- Claude hud hit 5 3k Github Stars Because Developers Were Flying Blind With Claude Code
- 27 Agents 109 Skills 88k Github Stars is Everything Claude Code Genius or Over Engineering
- Claw Code Rewrote Claude Code in Rust Before Sunrise and hit 50k Github Stars in 2 Hours
- From Claude Flow to Ruflo 22k Stars 5900 Commits and the Multi Agent Swarm Taking Over Claude Code
Top AI Product 
Leave a comment