BroadcastBrain: Archived TV to a Searchable Cable Library

    1

    Automate the segmentation, analysis, and organization of archived TV recordings into a searchable library.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    BroadcastBrain: Archived TV to a Searchable Cable Library

    Example session with this skill installed

    I have a 6-hour VHS rip of MTV from 1994 in my Incoming folder. How do I process it and find all the Pepsi commercials?

    • Read your context and instructions
    • Compiled the broadcastbrain

    I will initialize the processing pipeline for your MTV recording and then search the resulting library for the specific commercials.

    Processing and Search Results

    The recording has been indexed and segmented. Below is the summary of the bb run operation and the targeted search results for "Pepsi".

    Clip IDCategoryDurationConfidenceDetection Method
    3f2a91bccommercial00:300.98OCR: "Pepsi", Vision: Logo
    a07c5e12commercial00:300.94Speech: "Pepsi New Generation"
    c91d0b44commercial00:150.89OCR: "Pepsi"

    Processing Stats:

    • Source: mtv_1994_rip_01.mkv
    • Total Clips Found: 142
    • Shows: 4 (Beavis and Butt-Head, MTV News, etc.)
    • Commercials: 88
    • Unknown: 12 (Review with bb clips --category unknown)

    Next steps

    1. Run bb clip 3f2a91bc to see the full metadata and why it was categorized.
    2. Use bb dashboard to visually confirm the segments in your browser.
    3. Run bb resegment <source-id> --min-score 1.1 if you find clips are being split too frequently during music videos.

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

    About this skill

    Managing massive archives of VHS rips, DVR recordings, and DVD captures often results in a mess of unsorted files. BroadcastBrain solves this by automating the segmentation, deduplication, and categorization of old TV broadcasts into searchable libraries.

    What it does

    • Automatic segmentation splits long recordings into individual shows, commercials, promos, bumpers, and station IDs.
    • Multimodal analysis utilizes optional backends for Whisper-based speech-to-text, OCR for on-screen text, and vision models for logo detection.
    • Deduplication identifies and files duplicate clips using SHA-256, a perceptual hash and a Chromaprint audio fingerprint; the best copy stays in the library and alternates are filed separately, never deleted.
    • Plain-English search allows querying the database for specific brands, years, or content types like "all MTV clips from 2004."
    • Exporting generates schedule-ready data formatted specifically for FieldStation42 installations.

    How it works

    1. Initialize the environment by running bb init to create the directory tree and bb doctor to check for required binaries like ffmpeg.
    2. Process recordings by placing files in the Incoming/ folder and executing bb run or bb run --watch for overnight processing.
    3. Analyze and query the library using bb search or the bb dashboard to find specific commercials or clips.
    4. Refine results using bb resegment or bb recategorize to fix misclassified clips or adjust boundary sensitivity without re-running models.

    Frameworks & tools

    This tool is a Python-based CLI (bb) requiring Python 3.9+ and ffmpeg. It supports optional integrations with Whisper, Tesseract OCR, Chromaprint, and 7zip.

    Why this beats prompting it yourself

    Writing custom ffmpeg scripts to detect scene changes and commercials is error-prone and slow. This skill provides a specialized CLI interface that manages the database state, handles interruptions safely, and implements TV-specific classification logic that LLMs cannot perform on raw video bytes alone.

    Use cases

    • Convert bulk Internet Archive TV downloads into a categorized clip library.
    • Build a searchable database of retro commercials for creative reference.
    • Prepare a "FieldStation42" broadcast ready-set from raw VHS captures.
    • Find duplicate recordings across multiple drives: the best copy stays in the library, the rest are filed under Duplicates/ (nothing is ever deleted).

    Known limitations

    Requires ffmpeg and Python 3.9+. Deep analysis requires specific binary dependencies like Tesseract or Chromaprint for full functionality.

    Needs Python 3.9+ and ffmpeg. Each AI backend is an optional install: Whisper (speech), vision and embeddings are pip extras; OCR also needs the tesseract binary, and better duplicate detection needs chromaprint. Without them it still runs, with those analyzers marked skipped. A first run on hours of video can take all night.

    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

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 5 days ago

    Requires

    Git
    BroadcastBrain (free, github.com/hotdog-face/BroadcastBrain)
    Python 3.9+
    FFmpeg

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access
    Listed5 days ago

    What's inside

    Frequently Asked Questions

    BroadcastBrain automates the labor-intensive process of manually cutting and sorting long TV recordings. It uses multimodal analysis to identify where shows end and commercials begin, turning unsorted VHS or DVR rips into a searchable, categorized digital library.

    This skill is a Python-based CLI tool designed for local environments. It requires Python 3.9+ and ffmpeg installed on your system to perform video segmentation and library management.

    It identifies them: duplicates are matched by SHA-256, a perceptual hash sequence and a Chromaprint audio fingerprint. The best copy becomes the one in your library and the others are filed under Duplicates/, so browsing Commercials/ shows one copy of each spot. Nothing is deleted; alternates stay searchable.

    Once your library is processed, you can use the 'bb search' command or the dashboard to find content using natural language. For example, you can query for specific years, brands, or content types like '90s toy commercials' or 'MTV bumpers.'

    The skill: a SKILL.md that teaches your agent to install BroadcastBrain, pick optional AI backends from `bb doctor`, run it overnight, search it, and fix mistakes the cheap way (resegment, recategorize, reanalyze) instead of starting over. BroadcastBrain itself is free and MIT-licensed on GitHub (github.com/hotdog-face/BroadcastBrain); you install it separately.

    Absolutely. BroadcastBrain features an export function specifically formatted for FieldStation42 installations, making it easy to move your processed clips into a broadcast-ready schedule.

    New in Workflow Automation