Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    analyze video

    1

    Automated video analysis tool for transcription, OCR, and visual frame description via MCP.

    Free

    4 usesSecurity scanned
    analyze video

    analyze video

    Example session with this skill installed

    Analyze the video at https://www.youtube.com/watch?v=dQw4w9WgXcQ. I need the transcript and a description of what happens on screen.

    • Read your context and instructions
    • Compiled the analyze video

    I have completed the analysis of the provided video URL. The process involved transcribing the audio via Whisper and performing a visual analysis of sampled frames.

    Video Analysis Report: dQw4w9WgXcQ

    CategoryDetails
    Duration03:33
    LanguageEnglish (detected)
    StatusAnalysis Complete (Success)

    Transcription Summary

    The audio features a male vocalist performing a pop song. The lyrics center on themes of commitment and persistence. Key phrases detected include "Never gonna give you up," "Never gonna let you down," and "A full commitment's what I'm thinking of."

    Visual Description

    TimecodeVisual Content DescriptionOn-Screen Text
    00:00:15A man with reddish-brown hair in a trench coat dancing in an industrial setting.N/A
    00:01:05Close-up of the singer performing against a black background with bright stage lights.N/A
    00:02:10A bartender performing acrobatic flips behind a wooden bar counter.N/A
    00:03:00The singer dancing in a white outfit with backup dancers in a fenced urban area.N/A

    Technical Metadata

    • Tool Used: analyze_video
    • Frames Sampled: 12
    • Audio Provider: Groq (Whisper-large-v3)
    • Cache Status: Miss (New analysis performed)

    Next steps

    • Would you like me to generate a full Markdown transcript with timestamps?
    • I can extract specific frames as high-resolution images if needed.
    • Do you want a summary optimized for social media metadata?

    Connects securely to your tools. The creator never sees your data.

    About this skill

    The problem

    Manually transcribing videos or logging visual changes for accessibility and SEO is time consuming. Developers often struggle to integrate multiple disparate tools like Whisper, OCR, and vision models into a single automated pipeline.

    What it does

    • Transcribes audio content using Whisper via the Groq API for high-speed processing.
    • Extracts on-screen text and phrases using specialized OCR models.
    • Generates frame-by-frame visual descriptions to document scene changes and actions.
    • Handles video fetching automatically from YouTube, Vimeo, or direct URLs via yt-dlp.
    • Manages session caching to prevent redundant API calls and reduce costs.

    Frameworks & tools

    yt-dlp, Whisper (Groq), NVIDIA NIM, OpenAI Vision, and Python.

    Why this beats prompting it yourself

    Standard LLM prompting cannot "see" or "hear" a raw URL without a middleware bridge. This skill abstracts the complex plumbing of downloading, frame sampling, and multi-modal API orchestration into a single MCP tool call.

    Use cases

    • Generating detailed accessibility descriptions and alt-text for video libraries.
    • Creating searchable indexes of webinars by extracting slide text and transcripts.
    • Automating content moderation by auditing visual and verbal triggers in uploaded media.

    Known limitations

    Requires an active internet connection to reach upstream APIs. Local processing of Whisper requires pip installation of faster-whisper via the auto-install flag.

    Requires internet access for API calls. Performance for local transcription depends on system resources if auto-installing faster-whisper. Requires valid API keys for vision providers.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    4 uses

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Permissions

    No special permissions declared or detected

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    This skill acts as an MCP bridge that automates the downloading, frame extraction, and multi-modal analysis of videos, eliminating the need to manually upload files or stitch together separate transcription and vision APIs.

    The tool supports a wide range of sources including YouTube, Vimeo, and direct links to MP4 or MOV files, leveraging the yt-dlp library for reliable media fetching.

    Yes, the skill includes a session caching mechanism that stores results from previous analyses to prevent redundant API calls, helping you manage and reduce your processing costs.

    The skill is designed to work with any AI agent or platform that supports the Model Context Protocol (MCP), allowing it to function as a seamless extension of your existing AI workflow.

    While the skill uses high-speed cloud APIs for transcription by default, it includes an auto-install flag for 'faster-whisper' if you prefer to handle audio transcription via local processing.

    You will receive the full MCP server implementation, documentation for setup, and access to future version updates as new vision and audio models are integrated.