final-decisionlisted

# Final Decision Compiles the complete, evidence-based standing of a candidate and presents it to the human so **they** can approve or reject it. This Skill drafts nothing on the human's behalf and never implies its own summary or recommendation IS the decision. Full explanation of what each status means lives in `references/decision-framework.md`; the exact gate/state/ranking mechanics live in `references/decision-gates.md`. Read both before using this Skill. ## When to use this - Once a candidate reaches `HUMAN_REVIEW` (the natural end of the happy-path chain — see `decision-gates.md` SM-04). - Or at any earlier point, as an interim standing check, via `hunt rank` alone — this is safe to run repeatedly and has no side effects beyond recording the recomputed `final_status` (`decision-framework.md`). ## Step 1 — Get current standing ``` hunt rank <run_id> ``` This re-applies hard gates and Pareto ranking using only evidence already collected (fetches nothing new). It sets `final_status` to `REJECTED` (gate fired), `NEEDS_MORE_DATA` (gates passed, zero ranking dimensions computed), or `PROMISING` (gates passed, Pareto-ranked) — **never** `QUALIFIED`. Always re-run this rather than relying on a standing computed earlier in the conversation; new evidence may have changed it. ## Step 2 — Get the full evidence report ``` hunt show <opportunity_id> ``` This is the full evidence-first report. Every section without data says so explicitly (`NOT_YET_ANALYZED`, `NO_DATA`