social-video-analyzer

What This Skill Does

The social-video-analyzer is a powerful OpenClaw agent skill designed to bridge the gap between multimedia content and actionable intelligence. By leveraging yt-dlp for robust video acquisition and Google Gemini's native video understanding capabilities, this skill allows users to ingest, transcribe, and deeply analyze videos from over 1,000 social media platforms including YouTube, TikTok, Instagram, and X. Unlike traditional methods that rely on basic speech-to-text, this skill performs scene-by-scene visual analysis, extracts complex context, and generates structured summaries. It acts as a comprehensive multimedia research assistant that can handle tasks ranging from high-level sentiment analysis to granular production quality critiques.

Installation

To integrate this skill into your environment, use the OpenClaw command-line interface. Run the following command in your terminal to download and prepare the repository: clawhub install openclaw/skills/skills/aiwithabidi/a6-social-video-analyzer Ensure you have your GOOGLE_AI_API_KEY configured in your environment variables, as this is strictly required for the Gemini processing layer to function. Once installed, you can trigger the analysis via the provided Python or Shell scripts found within the skill's scripts directory.

Use Cases

This skill is highly versatile for professionals across multiple domains. Content creators can utilize it to repurpose long-form videos into bite-sized social media posts by identifying key highlights and quotes. Marketers can perform competitive intelligence by analyzing the hooks, calls-to-action (CTAs), and production styles of top-performing ads. Educational researchers can quickly index massive video archives for specific keywords or visual themes, while social media managers can use it for automated trend tracking and influencer content audits. It essentially transforms raw video files into structured, queryable data.

Example Prompts

1. "Analyze this video URL and tell me the exact timestamp where the creator mentions their main product, then draft three variations of a tweet summarizing that point."
2. "Watch this competitor ad and list the visual cues and marketing techniques used to keep viewers engaged for the first 5 seconds."
3. "Transcribe this tutorial and extract a checklist of every step mentioned in the video, ignoring the introduction and outro content."

Tips & Limitations

To maximize performance, ensure the videos you target are under the 2GB limit mandated by the Gemini API. While the skill supports a vast array of platforms, private or age-restricted videos that require authentication may fail to download due to the limitations of yt-dlp. For the best results, use clear, high-resolution videos, as native video understanding performs significantly better when visual details are crisp. Always remember to manage your API quota if you plan on processing long, continuous streams of video content to avoid service interruptions.