메뉴
Use 48-72h or more after publishing to measure deployed Slope Studio bets, score virality, test assumptions, and update the learned channel strategy.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Operator playbook for long-form story/audiobook illustration videos with CONSISTENT recurring characters. Use when the user wants to illustrate an existing story or audiobook text as a video, build/rebuild a character card from a reference-image folder (characters/<name>/), or debug character look drift between generated scenes.
Operator playbook for long-form story/audiobook illustration videos with consistent recurring characters. Use when the user wants to illustrate an existing story or audiobook text as a video, build a character card from a reference-image folder (characters/<name>/) or from the story text alone (no pictures — whole-story analysis locks a constant look), storyboard prose into scenes, or debug character look drift between generated images.
Operator playbook for the Slope Studio short-video pipeline. Use when the user wants to produce, generate, render, run, debug, observe, or publish a short vertical video (TikTok / YouTube Shorts) from an idea — or to run/inspect any single stage (script, visuals, clips, stitch, voice, save, publish). Covers the `studio` CLI end-to-end: provider choice, cost control, artifact inspection, and troubleshooting.
Umbrella orchestrator for a Slope Studio channel's viral-growth loop. Use when the user wants to RUN THE WHOLE ideate→deploy→measure→learn cycle, or doesn't know which step they need, or wants to READ channel state / PICK the next bet / WRITE a growth brief (journal · backlog · report all live here). It composes the per-step lego-block skills (marketing-ideate, marketing-deploy, marketing-measure-learn) and the hands-off driver (marketing-autopilot). For one creative step, invoke that step's skill directly.
Use 48–72h+ after publishing to CLOSE the loop: first MEASURE (fetch stats + comments, score each deployed bet's virality relative to the channel's own portfolio — deterministic), then LEARN (reflect on which pre-stated assumptions held vs were refuted, extract win/lose patterns, set the next direction + idea seeds — agent-driven). One lego-block of the ideate→deploy→measure→learn growth loop; feeds back into marketing-ideate.
Use SPECIFICALLY for hands-off SCHEDULING of the growth loop — the unattended driver that keeps a channel turning on a cron/interval with no operator in the seat. Each tick it asks the engine what's due (measure matured videos / reflect / refill the backlog / produce the next bet) and does that ONE action, handling the deferred-measurement timing automatically — then defers the actual step-level work to the per-step skills. Set it on a recurring schedule (the /loop or /schedule skill, or cron). For a single manual step, invoke that step's skill directly instead.
| name | marketing-measure-learn |
| description | Use 48-72h or more after publishing to measure deployed Slope Studio bets, score virality, test assumptions, and update the learned channel strategy. |
Read the canonical skill before doing this work:
../../../.agent-instructions/skills/marketing-measure-learn/SKILL.md
