AI Job Search

AI Job Search is an open-source application framework built around Claude Code skills. You load a profile from documents, a pasted CV, or a setup interview, then run portal search, fit ranking, and a drafter-reviewer pipeline that outputs LaTeX CVs and cover letters. Core fit and drafting logic is country-agnostic; the bundled portal CLIs start on the Danish market and are meant to be swapped or extended for local boards. Community forks adapt the same pattern for other agent CLIs.

Most AI job tools upload your history to a SaaS black box. This one stays on your disk as a forkable Claude Code workflow the author used to land a real AI engineer role — 69 applications, 20 interviews, one signed contract. Founders and operators who hire (or hire themselves) get a transparent, editable playbook instead of another résumé chatbot.

Clone into a private repo (public forks expose personal data), install portal CLIs with the Python helper, and run /setup inside Claude Code. /scrape searches matching portals and dedupes results; /rank batch-scores when the list is long; /apply on a URL evaluates fit, drafts tailored materials, and runs a reviewer pass before final LaTeX output. Optional ATS parse checks use pypdf or pdftotext when available.

Not a hosted job board or a generic coding agent. It is a full local career workflow encoded as agent skills — profile, scrape, score, draft, critique — that you own and fork, with a proven personal funnel behind the README.