A job search is a sales pipeline, so it gets pipeline tooling. One script reads plain markdown notes on companies and people and keeps three surfaces in sync from them, a kanban board, a dashboard with org charts and follow-ups, and news alerts stamped on each company.
Job-search trackers usually live in a spreadsheet nobody updates, separate from the notes where the thinking happens. I wanted the notes to be the database and the views to be derived.
Each company and each contact is a markdown note with structured metadata in its header: stage, owner, last touch, next step. The engine reads them all, validates them, and regenerates the kanban and the dashboard as static files. A separate radar feeds news signals into the same metadata, so a funding round or a new commercial hire appears on the company card without anyone typing it.