Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Video Studio

    1

    Make finished videos or edit your footage: reference-style Reels/TikTok edits, talking heads, product demos, ads and explainers, with captions, sound and an editable project.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Video Studio

    Video Studio

    Example session with this skill installed

    Edit raw.mp4 in the style of reference.mp4. Keep my voice and add captions. Vertical for Reels.

    • Read your context and instructions
    • Compiled the video studio

    A job folder containing renders/final.mp4 (1080×1920, 30 fps), renders/poster.jpg, captions.srt, plan.json, STORYBOARD.md, project/ (the Hyperframes composition) and qc.json listing the passed checks and anything left for human review.

    captions.srt, plan.json, STORYBOARD.md

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

    About this skill

    Turn raw footage or a product link into a finished video, plus an editable project you can keep working on.

    Video Studio gives your agent one clear workflow for video work: understand the inputs, write a timed plan, build the picture, captions and sound, render, check its own output, and deliver. You don't need a pile of separate skills that disagree with each other.

    What you can ask for

    • Edit in the style of a reference. "Cut my raw footage like this Reel. Keep my voice and add captions." The agent studies the reference's pacing, framing, type and transitions, then applies them to your footage.
    • Talking-head clean-up. It keeps the best takes, drops false starts and dead air, and keeps captions in sync after every cut.
    • Product demos and launch videos from a URL, PRD or screenshots. It uses your real product screens and claims your sources support, never invented features or testimonials.
    • Explainers, ads, slideshows, audiograms, recaps and more, with format recipes for each.
    • Revisions. Give feedback on an existing project and it fixes that project instead of starting over.

    What you get back

    • A final MP4 and a separate poster frame
    • An SRT caption file when there's speech
    • The editable project (Hyperframes or FFmpeg) and a plan.json timeline
    • A QC report saying what was checked and what wasn't

    Built to be honest

    • Your original voice is kept by default. AI voiceover via ElevenLabs is optional and only used when you ask for it.
    • It won't claim it watched a link it couldn't download. It asks you to upload the file instead.
    • It never invents prices, testimonials, metrics or deadlines.
    • Every render is checked for duration, frame rate, audio and full-file decoding before delivery, and it says when visual review still needs a human.

    Included tools

    video_tool.py inspects media (contact sheets, cut and silence candidates), validates plans, assembles rough cuts, exports captions and verifies renders. Also included: an optional ElevenLabs helper and offline smoke tests.

    Requirements

    Python 3.10+ and FFmpeg. For animated composition: Node 22+, Hyperframes 0.8.97 and a headless browser. An ElevenLabs key is optional.

    Built from open-source MIT and Apache-2.0 video skills, rewritten into one workflow. The license notices are included.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Social Media.

    Does not include speech-to-text, semantic video analysis, Instagram/TikTok downloading, face tracking, B-roll search or cloud hosting; these need your own configured providers. Remotion, Manim, Blender and WebGL adapters are optional and not installed. The bundled FFmpeg roughcut does base assembly only; styled edits are built in a compositor such as Hyperframes. Tested on synthetic media, not on real customer footage.

    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 buying a skill.

    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 21 days ago

    Software
    Python 3.10+FFmpegffprobefetching
    Optional
    Node.js 22+Hyperframes 0.8.97Headless browserElevenLabs API key

    Node.js, Hyperframes and the headless browser are for motion composition. The ElevenLabs key adds voiceover and sound effects, billed by ElevenLabs.

    Permissions

    Allowed Hosts

    api.elevenlabs.io (optional, only when voiceover or SFX is requested)

    File Scopes

    A
    job
    folder
    you
    choose
    kept
    outside
    the
    skill

    The ElevenLabs key is read from the ELEVENLABS_API_KEY environment variable and is never written to files or logs. No other network calls are made by the bundled scripts.

    Listed21 days ago

    What's inside

    Frequently Asked Questions

    No. It's only used if you ask for AI voiceover or generated sound effects. Your original audio is kept by default.

    Not by itself. If your agent has a download tool it will use it; otherwise it asks you to upload the reference video.

    Only for animated text, motion graphics and styled captions. Simple cuts and assembly work with FFmpeg alone.

    No. It only uses facts and screens from your sources, and asks when proof is missing.

    New in Social Media