discover-companieslisted

# discover-companies Discovers **companies** likely to be hiring the candidate's target roles, resolves each to a careers board the scanner can sweep, screens them, and proposes adding them to the CareerRat source config (`careerrat companies`: SQLite in DB workspaces, legacy `config/sourced-scan.json` otherwise). Never writes a company without explicit user confirmation. Never duplicates a company already tracked, applied to, sourced, or excluded. > **Runs under AGENTS.md.** These contracts bind without being restated here: Privacy Invariant (`current_base` never outbound), Honesty Firewall, Placeholder/Bracket Ban, Gate Write-back, Domain-Neutral Rule, Browser Automation Contract, Activity Pulse logging, Tracker verify+snapshot, and Sent-Clears-Draft. Inline reminders at point-of-use are intentional; standalone restatements point back to the relevant AGENTS.md section. This is the company analog of `research-boards`. `research-boards` finds **boards/aggregators**; `discover-companies` finds **employers** and wires their ATS board into the sweep. The sweep (`search-jobs` → `scan:sourced`) only ever scans configured tracked companies (`careerrat companies`: SQLite in DB workspaces, legacy `config/sourced-scan.json` otherwise) — it has no discovery of its own. This skill is what grows that list, so future sweeps reach beyond the set the candidate has already exhausted. The configured scan list is not the discovery universe. `company_preferences.examples` are priority seeds