More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    YouTube News Video

    1

    Researches the day's news, writes voice scripts, voices them, and builds a cinematic animated YouTube episode plus 2 Shorts and a thumbnail, all in code.

    $10

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    YouTube News Video

    Example session with this skill installed

    Make today's news episode: 4 world and tech stories, about 3 minutes.

    • Read your context and instructions
    • Compiled the youtube news video

    Phase 1: a doc with 4 verified stories and a source list, voice scripts saved as 00_intro.txt, 01_story.txt … 99_outro.txt, short_A.txt and short_B.txt, 3 title options, a description with chapters, tags, Shorts metadata and an image wish list, then questions on voice, images and a reference animation.

    Phase 2: voiced and Whisper-checked clips, a 1920×1080 episode with captions and a -14 LUFS mix, 2 Shorts at 1080×1920, a thumbnail, and a report of every quality check marked measured or eyeballed.

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

    About this skill

    A complete YouTube news episode pipeline in two phases, from a one-line request to finished MP4s.

    Phase 1: research and writing

    • Asks how long the episode should be (about 2, 5 or 8-12 minutes, or your own length).
    • Researches stories from the last 24-48 hours and confirms each against at least two reputable sources (or a wire plus the official source), politically neutral and in its own words.
    • Writes voice scripts with fixed clip names, the YouTube title, description, chapters, sources and tags, plus titles and hashtags for 2 Shorts.
    • Gives an image wish list per story (photos of real people are fine when you supply them) and asks whether you have a reference animation to match.

    Phase 2: voice and build

    • Three voice routes: ElevenLabs with your API key, clips you record yourself in ElevenLabs, or a free local voice (Kokoro, open source) with setup steps for Apple Silicon and Intel Macs, Windows and Linux. Every clip is checked against its script with Whisper.
    • A "Cinematic Desk" look built in Remotion and Three.js: printed news briefs under window-blind light, kinetic serif headlines with motion trails, keyword ticks on the spoken word, a lit globe, big numbers, quote cards, and paper-plane flights through giant 3D category letters between stories.
    • A built-in director: a 130-entry camera bible plus shot-list rules that keep every shot different and on the beat.
    • Sound design and a -14 LUFS mix, burned-in captions and SRT files.
    • Measured quality checks with bundled scripts: every frame scanned for blank, dark, flashing or frozen frames, beat contact sheets, a camera-parking check, palette and frame colour checks, and a voice-to-picture sync check on the final video.
    • Reference matching: send a clip and its motion is measured frame by frame (timing, easing, overshoot) and rebuilt in your brand until every move is within ±0.08 s.
    • Delivers the 1080p long-form video, 2 vertical Shorts (1080×1920) and a 1280×720 thumbnail.

    What you can build with it

    • Daily or weekly news briefings: 3-5 stories in 2-5 minutes
    • Topic roundups: tech news of the week, sports weekend recap, business headlines
    • Breaking-news Shorts: one story, vertical, under 60 seconds
    • A regional news channel with stories picked for one country or audience, in its own brand colours

    Good to know

    • Standalone: the build engine, 13 scripts and all reference docs ship inside the skill.
    • Brand placeholders (channel name, palette, category colours, sign-on/off) are filled in once and stay consistent across episodes.
    • Doesn't fetch photos, footage or logos from news sites; images come only from you.
    • Rendering needs a computer (Mac, Windows or Linux). Asks before installing anything and before spending ElevenLabs credits.

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

    Requires Node.js and Python. Network access needed for ElevenLabs/HuggingFace downloads. Full 3D requires package installation capabilities.

    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 today

    Accounts
    Hugging Face
    Software
    PythonremotionNode.jsFFmpegto
    Optional
    Kokoroespeak-ng (free voice)ElevenLabs API key

    Remotion is free for individuals and companies of up to 3 people. ElevenLabs voices in monetised videos need a paid ElevenLabs plan.

    Permissions

    Terminal / Shell
    Browser
    Read Files
    Write Files
    Network Access

    Allowed Hosts

    api.elevenlabs.io
    huggingface.co
    registry.npmjs.org
    pypi.org

    File Scopes

    project
    folder
    (scripts/
    audio/
    anim/
    renders/)
    keys.env
    (read
    only)

    Network: api.elevenlabs.io only for voice route 1 and optional sound effects (sends the clip text or sound prompt plus the user's own key from keys.env); huggingface.co for the one-time Kokoro and Whisper model downloads; npm and PyPI for package installs, with the user's OK. The key is read only from keys.env, sent only to ElevenLabs, never printed. No admin (root) commands. Full table in SKILL.md under 'Network access'.

    Listedtoday

    What's inside

    Frequently Asked Questions

    No. You can use the free Kokoro voice that runs on your own computer, or record the clips yourself. ElevenLabs with your API key gives the most natural voice.

    The skill itself: SKILL.md, a build manual and 6 reference guides (camera bible, director rules, hero checklists, living motion, reference matching, sub-frame blur) and 13 scripts for voice, sound effects and quality checks. Your agent creates the Remotion project for each channel from these.

    A Mac, Windows or Linux computer with Node.js 18+, Python 3.10-3.12 and FFmpeg, and an agent that can run shell commands and search the web, such as Claude Code. The skill asks before installing anything.

    No. Every frame is original graphics built in code, plus any images you upload yourself (photos of real people are fine when you have the right to use them). It never takes footage, photos or logos from news sites.

    Every clip is transcribed with Whisper for word timings, and all cuts, ticks and captions are placed on those words. After the mix, a sync check transcribes the finished video and compares every word with the timeline.

    Yes. Fill in the brand section once (name, palette, category colours, fonts, sign-on and sign-off) and every episode stays consistent. You can also send a reference animation and it will match that motion in your own look.

    New in Image & Video