Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Local Script-to-Edit Toolkit

    1

    Turn your script and existing media into an organized, captioned short edit.

    $4

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Local Script-to-Edit Toolkit

    Example session with this skill installed

    Using my two cleared original geometric container images, product-0.png and product-1.png, assemble a silent 4-second 640x360 edit at 24 fps. Show each image for 2 seconds. Leave the first shot uncaptioned; use “A simple shape. A clear message.” on the second. Keep the supplied aspect ratio with letterboxing. Give me the MP4, SRT, source ledger and technical audit. Do not call an external generation service.

    • Read your context and instructions
    • Compiled the local script-to-edit toolkit
    • Generated the document

    Reproduced local test

    A 4-second silent edit was actually rendered from two original synthetic container images. This is a local acceptance example, not a live Agensi buyer-session recording. No external image, video or voice generator was called.

    Delivered files

    • film.mp4: 640 × 360, 24 fps, 96 decoded frames, H.264/yuv420p, 4.000 seconds; silent AAC stereo track.
    • captions.srt: one caption, correctly numbered after the uncaptioned opening shot.
    • assembly-report.json: both source hashes, trim offsets, timing, media/audio choices and final MP4 hash.
    • audit.json: complete decode, duration, frame and audio measurements.

    Complete edit manifest

    {
      "title": "Original geometric container study",
      "language": "en",
      "rights_confirmed": true,
      "size": [640, 360],
      "fps": 24,
      "shots": [
        {"id":"opening","kind":"image","path":"product-0.png","duration":2,"caption":"","audio_mode":"none","audio_kind":"none"},
        {"id":"detail","kind":"image","path":"product-1.png","duration":2,"caption":"A simple shape. A clear message.","audio_mode":"none","audio_kind":"none"}
      ]
    }
    

    Captions

    1
    00:00:02,000 --> 00:00:04,000
    A simple shape. A clear message.
    

    Source ledger

    • opening: product-0.png, 0–2 seconds, source trim 0, no caption, deliberately silent. Source SHA256: ecedd9d52291315c4b7dfaf56cfbbbf0fadfed08713da0e6fdf39def134287ea.
    • detail: product-1.png, 2–4 seconds, source trim 0, one caption, deliberately silent. Source SHA256: 1c5a3834bf4bf6d22dfdd3fef2f625291d2dab27be7c9ca662ad2fc5be5e27fe.
    • MP4 SHA256: 069a72ed8732953841b2564942c5314df677fbc959a762cd991744b257ea2a41.

    Technical review

    Full decode passed; measured duration matched 4 seconds; 96 frames decoded. Audio signal peak and RMS were zero, as requested. Pixel-hash differences are descriptive only, not proof of motion. The inspected final frame retained letterboxing and readable captions. The preview image below is from this actual render.

    Next: review the full playback in your own player before delivery, check all rights and destination requirements, and keep the source ledger with the edit. This example does not establish voice quality, meaningful motion, commercial rights or platform acceptance. The rights_confirmed field records a declaration, not automated legal verification.

    product-0.png

    PDF · document

    Generated

    Example file from a real run - the skill writes it into your workspace.

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

    About this skill

    Turn your script and existing media into an organized, captioned short edit.

    This is a local supplied-media editing workflow, not a text-to-film generator. Prepare an English or Chinese shot breakdown, continuity notes and source ledger, then assemble explicitly supplied clips or stills with audio and burned-in captions. Four standalone Python helpers support local assembly, technical auditing, a fixed procedural demonstration and optional basic English scratch narration.

    Included: agent instructions, input/output contract, source code, original example scripts, dependency/rights notes and a cold-start guide. Best for storyboard animatics, product still sequences and repeatable short edits. Supports 1–30 shots, up to 180 seconds, hard cuts and letterboxing. Plans require assistant judgment; helpers do not parse arbitrary screenplays.

    Requirements: a code-capable assistant or terminal, Python 3.9+, Pillow, FFmpeg/ffprobe with libx264/AAC, and a separately licensed caption font. Dependencies and media are not bundled. No API keys or model credits are required for this local workflow. Host costs may apply.

    Not included: new live-action or AI scene generation, premium voiceover, natural Chinese speech, music mixing, automatic word alignment, character consistency guarantees, lip-sync, stock media licenses, hosting or publishing. Inputs and applicable codec/service permissions must be cleared separately; this is not legal clearance.

    Test evidence: fresh English/Chinese captioned geometric robot demos, a new silent still-image edit, standalone helper runs and a 30-shot/180-second small-frame boundary render passed local technical checks. Preview images show actual output samples. The fixed robot demo is procedural animation, not arbitrary AI film generation. Sampled frames were visually reviewed; no real-time playback/listening or live Agensi buyer-host execution was tested. Technical checks do not certify artistic, auditory, rights or platform quality.

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

    Supplied-media editing only. No new live-action/AI scenes, natural Chinese TTS, premium narration, automatic word alignment, music mixing, lip-sync or character-consistency guarantee. Hard cuts; 30 shots/180 seconds maximum. No real-time listening/full-playback review or live Agensi buyer execution tested.

    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 7 days ago

    Software
    Python 3.9+FFmpegPillow

    MP4 export needs a code-capable host, Python 3.9+, Pillow, FFmpeg/ffprobe with libx264/AAC and a licensed font. Dependencies, media rights and host costs are separate. Optional Flite makes only basic English scratch narration.

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    Reads only user-designated local manifests, media and fonts; writes a new chosen output folder. No network, account connection, image/video provider or automatic publishing. Existing outputs are refused. Rights declarations are not legal verification.

    Listed7 days ago

    What's inside

    Frequently Asked Questions

    No. It prepares a production packet and edits media you supply. It does not generate new scenes. The included procedural renderer produces only the fixed geometric robot demonstration. External scene generation, natural Chinese speech, lip-sync and character consistency are not included.

    A code-capable host with local file access, Python 3.9+, Pillow, FFmpeg/ffprobe with libx264/AAC and a licensed caption font. The bundle contains text and standalone source only. Dependencies, source media and host costs are separate. Text-only hosts can guide planning, but cannot render the MP4.

    Local regression and synthetic-media checks, fresh English/Chinese captioned demos, a silent still edit and a small-frame 30-shot/180-second boundary render passed. Samples were inspected as frames. No real-time playback/listening, natural Chinese speech, full AI scene generation, clean-machine install or live Agensi buyer-host execution was tested.

    New in Workflow Automation