740 Jobs Scored, 1 Dream Role: How Career-Ops Turns Claude Code Into an AI Job Hunter

Watch the 60-second breakdown on YouTube:

The story behind the number

Santiago Fernandez (@santifer on GitHub) was tired of the job-search grind that every laid-off engineer in 2024 and 2025 hit: read a posting, decide if it’s worth it, tailor a résumé, write a cover letter, prep for the interview, follow up, repeat. He built an open-source command-center called Career-Ops that wraps Claude Code (or any AI CLI like Codex, Gemini CLI, or Cursor agents) and runs the whole loop for him. He fed it 740 listings. It surfaced the one that mattered. He got the offer. The story got picked up by WIRED and Business Insider, and the repo has crossed 51,000 GitHub stars in weeks.

What Career-Ops actually does

Strip away the hype and Career-Ops is a thin but very opinionated orchestration layer that sits on top of a coding agent. It treats your job hunt the way a senior engineer treats a sprint:

• Ingestion — scrapes job boards (LinkedIn, Indeed, Greenhouse, Lever, Ashby, Workable, Remote OK, We Work Remotely) into a normalized JSON spec.
• Scoring — runs each listing through a customizable rubric: comp band, remote policy, tech stack, company stage, vibe. Listings under a threshold are auto-archived.
• Tailoring — for high-scoring roles, the agent rewrites your résumé and cover letter against the job’s exact keywords without lying.
• Outreach — drafts the application, the LinkedIn DM to the hiring manager, and the follow-up sequence.
• Interview prep — pulls the company’s recent news, GitHub activity, blog posts, and Glassdoor signals into a one-page brief.
• State — every interaction is a markdown file you own, sitting in a Git repo you control.

Install it in five minutes

The friction floor is very low if you already have Claude Code installed. From a fresh terminal:

git clone https://github.com/santifer/career-ops.git
cd career-ops
npm install
cp .env.example .env
# Add your ANTHROPIC_API_KEY (Claude Code uses your existing session if logged in)
npm run init

The init step asks for your target role, comp band, geography, and the boards you want monitored. It writes the config to ~/.career-ops/ and creates a Git repo for your application history.

The hidden cost nobody mentions

The repo is free and MIT-licensed. Claude Code on the Pro plan ($20/mo) is enough for casual use. The hidden cost is your reputation if you blast generic AI-written applications. Every recruiter on LinkedIn can spot a cover letter that was clearly generated, and the second they pattern-match they delete every other application from your name. Career-Ops gives you the leverage to apply to fewer roles deeper, not more roles shallower. Use the scoring threshold aggressively. The default is too generous.

Where it fits in the AI agent ecosystem

Career-Ops is the cleanest production example I’ve seen of the new pattern where a CLI agent (Claude Code, Codex, Gemini CLI, Cursor) is the runtime and your domain logic lives as prompt files in a Git repo. It is not a SaaS. It is not an MCP server. It is a personal workflow you own. That same pattern is going to eat:

• Sales outreach (lead scoring + drafting + follow-up)
• Investor diligence (deal sourcing + memo writing)
• Personal CRM (relationship maintenance)
• Content production (this very pipeline that produced this post)

Want to automate your own funnel like this?

If you are building a customer pipeline rather than a job pipeline, the closest equivalent for SMBs is GoHighLevel. Different domain — leads, not jobs — but the same idea: one source of truth, automated scoring, automated outreach, all the state in one place. Affiliate disclosure: that link is my affiliate link. If you start a trial through it I get a small commission at no cost to you.

Listen to the long-form breakdown

I cover this episode and the rest of the week’s AI-tooling drops on the GPT AI Clips Podcast. Subscribe on Spotify here — new episode every weekday.

Take it further

• Watch the full 60-second short: YouTube
• Subscribe to the channel: @gptaiclips on YouTube
• Browse every AI tooling deep-dive: gptaiclips.com

Career-Ops repo: github.com/santifer/career-ops · MIT licensed · 51k+ stars.