jobhuntlisted

# /jobhunt — Job Search Orchestrator (multi-candidate) Orchestrates the full job-search lifecycle: **source → match → apply → log → monitor → update** for **any candidate**. Per-candidate profiles live under `candidates/<slug>/`. The agent is generic — Joshua-specific content lives only in `candidates/joshua/profile.md`, not in this skill. ## Required Reading (every invocation) ``` 1. ~/.claude/skills/jobhunt/candidates/<slug>/profile.md — hard filters, target roles, skills, projects 2. ~/.claude/skills/jobhunt/candidates/<slug>/rubric.md — scoring weights (optional; uses default if missing) 3. /Users/whfjoshua/Job_Hunt/pipeline.json — current entries, pending_review (scoped by candidate_slug) 4. ~/.claude/skills/jobhunt/SOURCES.md — target company boards 5. ~/.claude/skills/jobhunt/cv-parser.md — for ingest-cv 6. ~/.claude/skills/jobhunt/cover-letter.md — for apply (template) 7. ~/.claude/skills/jobhunt/email-patterns.md — for poll ``` `<slug>` is determined by the **Candidate resolution** step below. ## Candidate resolution (runs at the top of every subcommand) ``` 1. If --candidate=<slug> flag was passed → use it. 2. Else if pipeline.json entries exist → use the most-recently-active candidate_slug. 3. Else if ~/.claude/skills/jobhunt/candidates/<slug>/profile.md exists for exactly one candidate → use it. 4. Else if multiple candidate folders exist → ask user: "Which candidate? [