Career-Ops is a free, open-source job search platform that runs entirely on your machine and connects to Anthropic’s Claude through the Claude Code CLI.
The system evaluates job offers with a structured A-F scoring framework across 10 weighted dimensions, generates ATS-optimized CVs tailored to each job description, scans 45+ pre-configured company career portals automatically, and processes multiple job listings in parallel batches.
It’s great for technical candidates, operators, consultants, and other professionals who already work comfortably in a terminal and want one local system for resumes, reports, tracking data, and company research.
Career-Ops stores its work in local project files, uses Playwright for browser tasks and PDF generation, and adds an optional Go terminal dashboard for a more visual pipeline view.
Features
- Runs a full auto-pipeline from a single pasted URL: evaluation report, tailored PDF, and tracker entry in one step.
- Scores each offer A-F across 10 weighted dimensions covering role match, compensation, level fit, and growth trajectory.
- Produces a 6-block evaluation per offer: role summary, CV match, level strategy, compensation research, personalization notes, and STAR+R interview prep.
- Accumulates an interview story bank across evaluations (build 5 to 10 master STAR+Reflection narratives reusable across any behavioral question).
- Generates ATS-optimized CV PDFs with keyword injection per job description, typeset in Space Grotesk and DM Sans.
- Scans 45+ pre-configured company career portals (Anthropic, OpenAI, ElevenLabs, Retool, n8n, and others) across Ashby, Greenhouse, Lever, and Wellfound.
- Processes 10 or more offers in parallel via sub-agents in batch mode.
- Provides a terminal dashboard built in Go with Bubble Tea: 6 filter tabs, 4 sort modes, grouped/flat view, lazy-loaded previews, and inline status editing.
- Includes negotiation scripts covering salary frameworks, geographic discount pushback, and competing-offer leverage.
- Runs automated pipeline integrity checks: deduplication, status normalization, merge, and health verification.
See It In Action
How to Use It
Prerequisites: Claude Code installed and configured, Node.js 18+, Go 1.21+ (optional, for the dashboard).
1. Clone the repository and install dependencies:
git clone https://github.com/santifer/career-ops.git
cd career-ops
npm install
npx playwright install chromium2. Set up your profile by copying the example config:
cp config/profile.example.yml config/profile.yml3. Edit config/profile.yml with your name, email, target roles, career narrative, and proof points. Create cv.md in the project root with your full CV in Markdown format. This file is the source of truth for every evaluation and generated PDF. Optionally create article-digest.md with portfolio highlights and proof points.
4. Configure the portal scanner:
cp templates/portals.example.yml portals.yml5. Edit portals.yml to set title_filter.positive keywords, add companies to tracked_companies, and customize search_queries for your preferred job boards.
6. Open Claude Code in the project directory:
claude7. Paste a job URL or description directly, and Career-Ops runs the full pipeline automatically. You can also ask Claude to customize the system, for example, changing archetypes, updating your profile from a pasted CV, or adding companies to portals.yml. Claude reads the same mode files it executes, so it understands exactly what to modify.
8. Verify your configuration at any point:
node cv-sync-check.mjs # Check configuration integrity
node verify-pipeline.mjs # Check pipeline health9. Build and run the optional Go dashboard:
cd dashboard
go build -o career-dashboard .
./career-dashboard --path ..10. All available commands:
| Command | Action |
|---|---|
/career-ops | Show all available commands |
/career-ops {paste a JD} | Full auto-pipeline: evaluate + PDF + tracker |
/career-ops scan | Scan pre-configured portals for new offers |
/career-ops pdf | Generate ATS-optimized CV PDF |
/career-ops batch | Batch evaluate multiple offers in parallel |
/career-ops tracker | View application pipeline status |
/career-ops apply | Fill application forms with AI assistance |
/career-ops pipeline | Process pending URLs in the queue |
/career-ops contacto | Generate LinkedIn outreach message |
/career-ops deep | Run deep research on a specific company |
/career-ops training | Evaluate a course or certification against your goals |
/career-ops project | Evaluate a portfolio project for fit and presentation |
Pros
- Keeps job evaluation, resume generation, scanning, and tracking in one local system.
- Covers more than resume rewriting and reaches into research, tracking, and interview prep.
- Supports batch processing for higher volume searches.
- Keeps the final application decision manual.
Cons
- Requires Claude Code, terminal comfort, and local setup work.
- Needs strong profile context before evaluations become useful.
- Fits technical users better than casual job seekers.
Related Resources
- Career-Ops Case Study: The author’s walkthrough of the real job search that produced 740+ evaluated offers and one Head of Applied AI placement.
- Claude Code Documentation: Official setup and usage guide for the Claude Code CLI.
- Playwright Documentation: Reference for the browser automation layer Career-Ops uses for portal scanning and PDF generation.
FAQs
Q: Does Career-Ops submit applications automatically?
A: No. The system evaluates offers and fills in application form fields on request, but it never submits anything. Every submission requires a deliberate action from the user.
Q: How does the scoring system work?
A: Each job offer gets an A-F score calculated across 10 weighted dimensions. The system reads your cv.md and evaluates fit through reasoning about your experience against the job description, not keyword matching. The threshold for further consideration is 4.0 out of 5; below that, the system recommends not applying.
Q: Is my CV data sent to any server?
A: Your CV, profile, and evaluation data stay on your machine. The only external transmission is the API call to your chosen AI provider (Anthropic by default). Career-Ops does not collect, store, or transmit personal data to any third-party service.
Q: How accurate are the salary and compensation estimates?
A: The compensation research block pulls publicly available data and applies it to the role and location in context. The output is a research-backed estimate. The negotiation scripts are frameworks derived from common negotiation scenarios.
Q: Can Career-Ops be customized for non-AI job roles?
A: Yes, but it requires deliberate setup. The default archetypes (LLMOps, Agentic, PM, Solutions Architect) are optimized for tech and AI-sector roles. Users targeting other industries need to rewrite the archetype definitions and update the portal list.
Q: What happens if a portal changes its structure and breaks the scanner?
A: Portal scraping relies on Playwright and the Greenhouse API. If a company updates its career page structure, the scan may fail or return incomplete results.
