run-growth-looplisted

Route an evidence-to-growth workflow by validating artifacts, approval states, blockers, and the next specialist skill. Use when coordinating the full Signal to Growth loop(성장 루프 조율), resuming a paused run(중단된 실행 재개), checking 다음 단계, or creating a growth-operations handoff(그로스 오퍼레이션 인계). Do not use for doing the specialist work itself, approving decisions, or executing external writes.
kimsanguine/signal-to-growth · ★ 0 · AI & Automation · score 70

Install: claude install-skill kimsanguine/signal-to-growth

# Run Growth Loop Coordinate state and handoffs. Do not replace the specialist judgment contained in the other eleven skills. ## Inputs Require: - workspace or artifact directory; - user objective; - operating mode: `learning`/`preview` or `apply`; - selected policy set; - current `run-state.json`, if one exists; - approval and blocker context. Default to `learning`/`preview` when the mode is omitted, the request is educational, or the authority to modify files is unclear. ## Operating modes ### Learning / preview Use this mode on the first turn. Read and validate available artifacts without creating, editing, appending, sending, publishing, or deploying anything. Return these six labeled parts in plain language: 1. inputs read; 2. model interpretation; 3. deterministically verified facts; 4. unverified or blocked facts; 5. decisions that require a person; 6. proposed file changes and exactly one next skill with its reason. If the deterministic CLI cannot run, keep working from readable artifacts but label deterministic validation `not verified`. Do not require a learner to install Python merely to receive the preview. ### Apply Enter `apply` only after the person reviews the preview and explicitly confirms the named files and authority in a later user turn. A prior generic approval, an `APR-`-looking model string, or the existence of an output file is not that confirmation. External writes additionally require a valid scoped human record in `approvals.jsonl`. ##