job-huntlisted
Install: claude install-skill WorkflowtechAI/job-hunt
# Job Hunt
A search loop you can run every week. The person owns the decisions; this skill
does the sourcing, the honest scoring, and the paperwork.
Two tracks run at the same time off one identity:
- **Track A, employment.** Applications to jobs someone posted.
- **Track B, contract.** Engagements you propose to a company that never posted
anything.
Most people run only Track A and wonder why the pipeline is thin. The same
profile sells both ways.
## What this runs on
A folder of plain files. Default `~/job-hunt/`, anywhere works.
| File | What it is | Written by |
|---|---|---|
| `profile.md` | The candidate profile: identity, target roles, hard limits, honest claims | The intake, once. Edited whenever reality changes |
| `resume/master.md` | The master résumé as markdown: everything, in one file | The intake, once. Edited when a role or a number changes |
| `forms.md` | The field-shaped answers every application form asks for: dates, years per skill, authorization, availability | The intake, once. A line added after any form that asked something new |
| `answers.md` | The reusable spoken and free-text answers, in two registers | Built on request, rewritten as they get better at telling them |
| `runs/<date>.json` | One search run: the scored shortlist | Each search |
| `applications/<date>-<company>-<role>/` | One application: the posting, the review, the tailored résumé, the letter brief, the change log | Each apply pack |
| `pipeline.json` | Application status p