material-report

What This Skill Does

The material-report skill is an advanced creative intelligence tool for OpenClaw AI designed to deconstruct, analyze, and optimize video assets for performance advertising. By processing video files, the skill performs a granular audit of narrative structures—identifying hooks, problem-solution arcs, and call-to-action effectiveness. It identifies specific visual traits, tempo, and overlay densities to understand why a piece of creative succeeds. Beyond mere analysis, it acts as a creative strategist, generating entirely new video frameworks and providing detailed, time-stamped storyboards ready for production teams to execute. It effectively turns a single creative asset into a scalable testing pipeline.

Installation

To install this skill, use the ClawHub command-line tool: clawhub install openclaw/skills/skills/fevolq/material-report

Use Cases

1. Creative Scaling: Use this when you have a high-performing ad and want to generate three variations that maintain the winning hook but iterate on the body copy or CTA to combat ad fatigue.
2. Competitive Intelligence: Upload a competitor's video to dissect their visual language, keyword targeting, and structural pacing to inform your next campaign strategy.
3. Creative Refresh: Feed an old or underperforming video into the agent to receive a structural overhaul proposal based on current platform-specific trends and best practices.

Example Prompts

1. "Analyze this video file: [path/to/ad.mp4]. It's a mobile game ad for the RPG category on TikTok. Please extract the hook, map out the current structure, and propose a new 30-second storyboard with a more aggressive CTA."
2. "I'm running a SaaS product campaign on Meta. Review this video asset, summarize the key visual traits, and suggest a new framework that leans into social proof rather than the current feature-heavy approach."
3. "Here is my video. Please generate a detailed storyboard for a new version that targets a younger demographic, keeping the same 15-second duration but updating the tempo and on-screen text overlays."

Tips & Limitations

• ffmpeg Dependencies: The skill relies on ffmpeg for frame extraction. Ensure it is installed on your local system to avoid processing errors. If not available, provide a pre-extracted folder of frames.
• File Management: The skill generates a folder named after the video for frame storage. Do not modify these output paths, as the system needs them to correlate metadata during analysis.
• Context is Key: Always provide the category or vertical of your product. While the skill can attempt to infer it, providing the correct vertical ensures that the acquisition keywords and suggested frameworks align with your specific market's best practices.
• Iterative Workflow: If the initial analysis doesn't quite capture the nuance you are looking for, follow up by asking the agent to refine specific segments of the storyboard or to swap specific aesthetic styles.