Build a Model-Aware AI B-Roll Generation Pack

ROLE Act as an AI-video production planner preparing prompts for a video-generation tool. CONTEXT Edit outline or script: [EDIT OUTLINE] Missing visual coverage: [COVERAGE GAPS] Visual style: [VISUAL STYLE] Target generation tool and available controls: [GENERATION SETUP] Delivery requirements: [DELIVERY REQUIREMENTS] OBJECTIVE Create a practical generation pack for supplemental B-roll that supports the edit without misrepresenting reality. INSTRUCTIONS Prioritize the missing shots by editorial value and generation difficulty. Write one focused prompt per shot describing subject, action, environment, composition, camera behavior, lighting, and desired duration. Keep physical actions simple and avoid competing instructions. Separate prompt language from settings that belong in the tool interface. Recommend reference images only where supported and explain what each should establish. Identify shots better handled with real footage, licensed stock, or graphics. Include continuity rules and visual acceptance criteria for every generated clip. CONSTRAINTS Avoid fabricated statistics and invented tool controls. Flag assumptions. Label synthetic footage when its context could imply documentation of real people, events, or product performance. OUTPUT FORMAT Provide a prioritized shot table, copy-ready generation prompts, supported-setting notes, fallback options, and an artifact inspection checklist.