Career Ops: Free Open-Source AI Agent That Scores Every Job Out of 5 Before You Apply (Setup Guide)

By Mulazim TeamUpdated October 20266 min read
Career Ops: Apply Where You Will Actually Get Picked
76K+
GitHub stars
(growing daily)
1-5
Score for every job
against your CV
4.0
Below this it says
do not apply
MIT
Free, open source,
runs on your machine

What Happens After You Paste One Job
StepWhat Career Ops does
1Checks the posting is still open (not a ghost job or repost)
2Scores the role 1-5 against your real CV
3Under 4? Tells you to skip and saves your night
4Worth it? Drafts a tailored CV, cover letter and answers
5You review everything. You press Submit. Never the bot

Everyone sends the same CV to 200 places and waits. Career Ops flips that: it is a free, open-source AI job search agent that reads a job posting, checks it is still open, scores the fit against your actual CV out of 5, and tells you straight when a job is not worth your night. For the jobs that are worth it, it drafts a fresh CV tailored to that recruiter, plus a cover letter and application answers. You always press Submit yourself.

It runs locally on your own machine inside the AI CLI you already use (Claude Code, Codex, OpenCode and more). No account, no uploads, no telemetry. It was built by Santiago Fernández de Valderrama Aparicio (@santifer) after his own search: 740 listings evaluated, 68 applications, 12 interviews, 1 offer. He got the job, then open-sourced the filter. The repo has crossed 76,000 GitHub stars and was GitHub’s #1 trending repo.

Source 1 - Career Ops README (GitHub) github.com/career-ops-hq/career-ops
Install command, features, what it will never do · checked Oct 23, 2026
Source 2 - The author’s full story santifer.io/career-ops-system
How the first user (the author) ran his own search · checked Oct 23, 2026

How Career Ops Works

  1. Paste a job. A link or the full text, from any portal.
  2. Still-open check. It verifies the posting is live before you write a word, and flags reposts and possible ghost jobs.
  3. Fit score out of 5. It reads the JD against your real CV, requirement by requirement, with how much each one matters for that posting. Below 4.0 it recommends skipping. You can override it.
  4. Tailored drafts. For a job worth chasing it writes an ATS-friendly CV for that specific recruiter, a research-backed cover letter, and draft answers to the application fields.
  5. You decide. It never auto-submits, never sends an email, never contacts anyone by itself. Every draft waits for your review.

Everything stays on your machine: applications, tracker, stats. After a run of rejections it also names the skill gap, so the next application is smarter than the last.

Setup in 2 Steps

Step 1 - Install with one command. In your terminal run:

npx @santifer/career-ops init

npx ships with Node.js - it runs the installer once without installing anything globally. Check you have it with node -v; no Node yet? Install the LTS version from nodejs.org first. The installer clones the latest release into a career-ops folder and installs dependencies.

Step 2 - Open your AI CLI in that folder and say hello.

cd career-ops claude # or codex / opencode / qwen / agy / grok

On first launch Career Ops interviews you in plain chat: your CV, what roles you want, what you refuse. That conversation is the setup prompt - nothing to edit by hand, nothing to configure. After that, paste any job and it starts scoring.

Optional - let your AI assistant run it for you. If you use a personal AI assistant (like the one from our reel), you can have it operate Career Ops on your machine for you: sign in only through the secure sign-in link the assistant itself gives you - never send a password or login cookie to anyone in chat - then tell it in one message: “Run Career Ops: here is my CV, score every job I send you out of 5, and draft a tailored CV for anything 4 or above.” It takes it from there.

What It Will Never Do

What It Costs

Career Ops itself is free and MIT licensed. The AI usage is yours: it runs in your own AI CLI, on your own subscription or API key. A free and local model path is documented in the repo (docs/RUNNING_ON_A_BUDGET.md) if you do not want to spend anything at all.

The Fine Print (Read This Before You Trust It)

Want the Repo Link in Your DMs?

Comment JOB on our latest Instagram reel (@amanarya.ai) and the repo link lands in your DMs automatically. Bookmark this guide for the full setup, and star the repo if it helps - it helps the next person find it.