Livestream Scene Designer — 直播空间造景

    0

    Reference-based, people-free livestream sets with camera, layout and material checks. Requires your own image-capable agent.

    $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 CLIDeepSeekDeepSeekMuseMuse+22 more

    Livestream Scene Designer — 直播空间造景

    Example session with this skill installed

    万圣节马上到了,从万圣节的各个主题出发找参考做50个万圣节直播空间场景。5个一组的相似风格。要求完全无人、真实相机光线和自然景深,避免地面过长及密集脏纹理。

    • Read your context and instructions
    • Compiled the livestream scene designer

    Historical real-session example — Halloween livestream sets

    The actual request was for 50 Halloween livestream scenes in 10 groups of five. Themes included pumpkins, witch apothecaries, Gothic interiors, cream decorations, candy shops, forests, vintage circus, celestial salons, parties and laboratories.

    The session produced 50 designs and 141 retained PNG versions with 141 matching Chinese descriptions. These historical totals are not a promised package size or recommended budget. Most images were resized to 1080 × 1920 from smaller native sources; one failed landscape version was retained as history.

    Review found excessive foreground floor, dense texture, oily highlights and insufficient optical depth. Selected later versions 05 (3), 09 (3) and 18 (3) improved near-use-plane focus and softer rear scenery. Some forest, circus and party outputs still looked theatrical; scene 39 still had inadequate sofa crop margin. The batch was not declared universally successful.

    The subsequent skill revision added pre-generation camera/material/light planning and representative-scene batch gates. Format and file-workflow checks passed, but improved first-pass success is not yet guaranteed by a new image benchmark. No extra images were generated solely for this listing.

    Example companion description

    完全无人、零人物,9:16竖幅,正面水平机位。乡村秋季烘焙直播空间,近前景木操作台两端放南瓜派、陶碗、木勺、鸡蛋和陶壶,中央台面留空,下层收纳碗器、南瓜与织篮。左后壁炉、后墙灰绿矮柜与黑框窗、右侧落地层架形成不对称布景,秋叶和南瓜在侧后成组。近台面和器皿自然清楚,远柜与窗景逐渐柔化,柔和窗光与局部暖灯融合。无人物、人影、人体局部、文字或水印。

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

    About this skill

    Design complete, people-free livestream sets

    Turn reference images and product material into complete livestream spaces for standing presentation, seated conversation or tabletop demonstrations. Organize architecture, props, displays and adult-scale empty use areas with credible supports and access paths.

    First-pass planning and actual image review

    Before generating, check camera height, framing, excessive foreground floor, prop density, texture scale, main light, neutral color and focus distance. Batch jobs validate a planned representative scene before expanding similar designs. Recurring defects pause affected generation instead of repeating them across the batch.

    The default two-round workflow uses reference images, actual visual inspection and targeted refinement. It checks people-free output, visible construction, use space, crop safety, material and optical depth. Model outputs vary: photorealism, flawless construction and first-pass success are not guaranteed. Agent visual review is not engineering certification or an automatic quality guarantee.

    Delivery and requirements

    Actual PNG versions plus matching standalone UTF-8 Chinese scene descriptions. Default target: portrait 1080 × 1920; source dimensions and resizing are disclosed. Requires an image-capable host with reference-image input, image viewing, supporting-file access and local file output. Python 3.10+ and Pillow are required for delivery scripts. A text-only agent cannot deliver images. Agensi access does not include image tools, credits or host subscriptions. Instructions and companion descriptions are primarily Chinese. Best suited to image-capable ChatGPT/Codex; other agents are not promised equivalent results.

    US$5 once — future updates included

    One-time purchase, no recurring subscription. Existing buyers receive future published updates to this same skill without paying for the skill again. New versions will be submitted to this listing, subject to platform review and availability. No fixed update schedule is promised. Image credits, third-party subscriptions, custom scene production and consulting are not included.

    Open-source transparency

    This is the creator's own MIT-licensed project, also available free at https://github.com/SJ3199/livestream-scene-skill . Buying through Agensi is optional support for development and marketplace access, not an exclusive paid edition. MIT license rights remain unchanged.

    中文

    US$5一次买断,无订阅;已购用户免费获得本技能后续发布更新。需要自备支持生图、参考图输入、看图和文件读写的宿主及额度。GitHub保留MIT开源免费版。后续轮次只处理实际输出偏差,不保证每张首稿均达标。

    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 an image-capable host, reference-image input, image inspection, bundled-file access and local read/write tools; Python 3.10+ and Pillow for scripts. Chinese-first instructions. Does not include image credits, a model or consulting. Visual quality varies; no engineering certification or guaranteed first-pass success. Precise logo/text replication is not guaranteed.

    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 25 days ago

    Software
    ChatGPT
    orCodex
    Python 3.10+Pillow

    ChatGPT or Codex must support image inspection and reference-image generation and editing.

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    Read user-provided reference/product/logo assets and skill files; write PNG/TXT versions to the selected output folder. Image processing uses the host image service and credits. No credential collection or unrelated file access. Visual review is not engineering certification.

    Listed25 days ago

    What's inside

    Frequently Asked Questions

    Yes. US$5 once, no recurring subscription. Existing buyers receive future published updates to this same skill without another skill purchase, subject to platform review and availability. No fixed update schedule is promised.

    Your host must generate and edit images from references, inspect images, read supporting files and write outputs. Python 3.10+ with Pillow runs the helper scripts. Image service access and credits are your own; a text-only agent cannot deliver images.

    Yes. The creator also publishes this MIT-licensed project free on GitHub: https://github.com/SJ3199/livestream-scene-skill . This listing is an optional support and marketplace-access channel, not an exclusive paid edition. MIT rights remain unchanged.

    No. The agent actually reviews output images for visible supports, scale, use space, cropping, material and depth. File scripts check naming, pairing and dimensions. Neither is engineering certification or an automatic guarantee of image quality.

    Yes, provide authorized scene references, product images and logos with the intended standing, seated or tabletop use. The workflow separates their roles and avoids copying unrelated branding. Exact text/logo reproduction depends on the host tools and is not guaranteed.

    New in Image & Video