Turning the AI resume screen back on the companies

What it does

Career-Ops plugs into AI coding CLIs like Claude Code and Gemini CLI to run a structured job-search pipeline. You paste a job URL or description; it classifies the role archetype, runs an A-F evaluation across ten weighted dimensions against your cv.md, and produces a scored report, an ATS-optimized PDF, and a tracker entry. It can also scan 45+ pre-configured company portals on major boards like Greenhouse and Ashby, batch-process listings in parallel, and accumulate an interview story bank of STAR+Reflection narratives over time.

The interesting bit

The system is explicitly designed as a filter, not a blaster: it recommends against applying to anything scoring below 4.0/5 and never submits an application without your sign-off. The author used it to evaluate 740+ offers and land a Head of Applied AI role, which suggests the rigor is in the scoring and pipeline integrity checks, not brute-force volume.

Key highlights

• Structured A-F evaluation across ten weighted dimensions, plus a six-block report covering role summary, CV match, and interview prep
• ATS-optimized PDF generation that injects keywords and customizes design per job description
• Portal scanner covering 45+ companies across Ashby, Greenhouse, Lever, Wellfound, and others
• Interview story bank that accumulates STAR+Reflection narratives across evaluations
• Human-in-the-loop design: AI recommends, you decide; it will not auto-submit applications

Caveats

• Early evaluations are weak until you feed it substantial context (your CV, career story, preferences, proof points); the README compares this to onboarding a new recruiter who needs a week to learn you
• Stale job postings can leak into the pipeline because some ATS feeds keep closed roles live; you need to pass --verify to launch a Playwright pass that drops expired entries
• The README warns this is not a “spray-and-pray” tool; if you are looking for volume application automation, this system will actively discourage you

Verdict

Worth a look if you are a developer targeting specific roles and want a disciplined evaluation pipeline that tracks everything. Skip it if you want a fire-and-forget application blaster—this system is explicitly built to reduce volume, not increase it.