company-healthlisted

# company-health Use this skill when the user asks how risky, stable, or healthy a company is — layoffs, financials, morale, "is this a safe place to land" — or when they ask to factor company risk into a role. Also use it **automatically** when a tracked company reaches the configured firing stage (default: the interview band), subject to the cost gate in STEP 1. The question is never "is this a good company." It is **"is *my function* a risky place to land at this company *right now*."** Role-scoped, time-stamped, judgment surface — a healthy company that just gutted the function you'd join is a red flag; a wobbly company doubling down on it may still be worth a shot. > **Runs under AGENTS.md.** These contracts bind without being restated here: Honesty > Firewall, Privacy Invariant (`current_base` never outbound), Placeholder/Bracket Ban, > Domain-Neutral Rule, Browser Automation Contract, Activity Pulse logging, and the Tracker > Write Contract (stamp → verify → snapshot → activity). Inline reminders at point-of-use > are intentional. **One extra hard rail for this skill: the rating is an INTERNAL signal > only — it must never appear in any outbound artifact** (cover letter, recruiter reply, > LinkedIn). No "your company is in freefall" energy reaches a human at the company. > **Agent voice.** Read `candidate/modes.yml#agent_voice` (default `standard`) before > presenting the rating. Apply the register from AGENTS.md#mode-switches. The persisted > `companyHealth` objec