Santiago has released career-ops, an open-source job search automation platform that transforms application management from manual spreadsheet tracking into an AI-powered pipeline. The system, built on Claude Code and released April 4, 2026, has garnered approximately 2.8K GitHub stars and 411 forks.
Single Command Interface Powers 14 Specialized Modes
Career-ops operates through a single slash command interface with 14 distinct skill modes. Users paste job descriptions or URLs, triggering an automated pipeline that produces evaluation reports with structured A-F scoring across 10 weighted dimensions, ATS-optimized PDFs with keyword injection, tracker entries in markdown tables, interview story banks in STAR+Reflection format, and salary negotiation frameworks.
The system combines multiple technologies: Claude Code for the agent layer with custom skills, Playwright/Puppeteer for PDF rendering with HTML templating, browser automation plus Greenhouse API integration for portal automation, a Go-based terminal UI using Bubble Tea framework with Catppuccin theming, and markdown tables with YAML configuration for data persistence.
Automated Portal Scanning and Batch Processing
Key features include automated portal scanning across 45+ pre-configured companies, parallel batch processing for evaluating multiple offers, dynamic PDF generation with keyword optimization, and a self-customizing system where Claude reads and edits the same configuration files it uses. Users can request changes through natural language, such as "Change archetypes to backend roles" or "Translate modes to English."
Santiago emphasizes applying only where genuine fit exists rather than mass-application tactics. He built this system after evaluating 740+ job offers and generating 100+ tailored CVs, ultimately landing a Head of Applied AI position.
Technical Architecture Emphasizes Modularity
The technical stack includes 14 skill modes covering evaluation, PDF generation, scanning, batch processing, and interview prep—each customizable through Claude's interactive interface. The Go dashboard provides a terminal-based interface, while markdown tables and YAML files handle data persistence. The system's self-customizing capability sets it apart from traditional job search tools.
Key Takeaways
- Career-ops has reached 2.8K GitHub stars and 411 forks since its April 4, 2026 release with MIT licensing
- The system evaluated 740+ job offers and generated 100+ tailored CVs for its creator, who landed a Head of Applied AI position
- Features 14 distinct skill modes including automated portal scanning across 45+ companies and parallel batch processing
- Built on Claude Code with a Go-based terminal UI, Playwright/Puppeteer for PDF rendering, and markdown/YAML for data persistence
- Self-customizing system allows users to modify configuration through natural language commands to Claude