The Problem
Job seekers face manual, time-consuming evaluation of offers, CV tailoring, and application tracking. Spreadsheets become unwieldy, and keyword-matching tools produce generic results that don't reflect actual fit or career positioning. The volume of modern postings makes systematic evaluation impractical without dedicated infrastructure.
What This Does
Career-ops is an AI-powered job search pipeline built on Claude Code that transforms the process into a structured command center. The modes/ directory contains 14 skill modes (apply.md, pipeline.md, batch.md, deep.md) that encapsulate distinct workflows—from single-offer evaluation to batch processing of 10+ offers in parallel using claude -p workers. The dashboard/ directory provides a Go-based TUI for browsing, filtering, and sorting the pipeline, with dashboard/main.go as the entry point. PDF generation leverages templates/cv-template.html and fonts in fonts/, injecting job-specific keywords for ATS optimization. A portal scanner (batch/batch-runner.sh) automates applications across 45+ pre-configured companies including Greenhouse, Ashby, and Lever. The batch/ directory enables parallel evaluation with sub-agents, while dashboard/internal/ui/screens/pipeline.go and dashboard/internal/ui/screens/viewer.go handle the terminal UI state. An integrity layer in modes/_shared.md and tracking files ensures merge safety, dedup, and status normalization across the pipeline.
How To Use It
Setup: git clone https://github.com/santifer/career-ops.git cd career-ops && npm install npx playwright install chromium
Validate prerequisites
npm run doctor
Configure: cp config/profile.example.yml config/profile.yml # Edit with CV details, preferences, proof points cp templates/cv-template.html ./ # Optional: customize PDF output
Running: The entry point is dashboard/main.go — start the Go dashboard server. For immediate evaluation, paste a job URL into the auto-pipeline mode, or use modes/auto-pipeline.md for URL-to-evaluation+PDF+tracker-entry workflow. Feed the system context: your CV, career story, preferences, and proof points in config/profile.yml. The system evaluates using a 6-block framework (role summary, CV match, level strategy, comp research, personalization, interview prep) and generates tailored PDFs. Review recommendations; the system never submits applications without your final call.
Real-World Use
A senior AI engineer evaluating 50+ offers uses the batch mode: claude -p --mode batch --input offers.yaml triggers parallel evaluation across sub-agents, producing scored entries (A-F across 10 weighted dimensions) and tailored PDFs in parallel. The interview story bank (interview-prep/story-bank.md) accumulates STAR+Reflection stories across evaluations, creating 5-10 master stories that answer any behavioral question. The negotiation scripts module provides salary frameworks and competing offer leverage. The dashboard TUI allows filtering by score, status, and company, with integrity checks ensuring no duplicate entries and normalized statuses across the pipeline.
Code Health & Issues
SDLC gap: No CI/CD pipeline detected—/.github/ contains only funding, issue templates, and a gitignore. No automated build/test gate exists. Test coverage: Only 2 test files found; no evidence of automated test execution in the workflow. Dependency hygiene: dashboard/go.mod and package.json are present, but no lockfile audit or vulnerability scanning configured. Configuration risk: config/profile.example.yml is committed as a template; users must copy and edit, but no .env or secrets management pattern is evident. Documentation strength: 10 doc files present, including docs/ARCHITECTURE.md, docs/CUSTOMIZATION.md, and docs/SETUP.md—well-maintained for onboarding. Structural organization: Clear separation between modes/ (skill definitions), dashboard/ (Go UI), batch/ (processing), and examples/ (worked cases).
The Bottom Line
This is a capable, well-structured AI job search system that excels as a filter and evaluator rather than a spray tool. The Go dashboard, batch processing, and PDF generation are functional and thoughtfully implemented. It's best suited for individual job seekers or small teams willing to invest time in onboarding the system with their specific context—CV, preferences, proof points. Organizations requiring out-of-the-box enterprise readiness will need to address the missing CI/CD and test infrastructure.