AGENT SKILL

direct-video-creation

Turn creative intent into a production-ready AI video brief, treatment, script/beat plan, shot cards, animatic/keyframe plan, node workflow, and model-profile prompts. Use for text-to-video, image-to-video, multi-shot or dialogue scenes, reference-video transformation, video editing or continuation, product/logo/jewelry/fashion/social/UGC films, fight scenes, character stories, cooking tutorials, music videos, award or freeze effects, talking-character clips, drone or one-shot footage, long-form-to-short clipping plans, storyboards, prompt audits, camera/light/sound direction, race/chase/kinetic montage, or improving a video-generation node.

License 许可
Apache-2.0
最近更新
2026-11-14
Stars
★ 299
Forks
50

01 前言 Front Matter

name
direct-video-creation
description
Turn creative intent into a production-ready AI video brief, treatment, script/beat plan, shot cards, animatic/keyframe plan, node workflow, and model-profile prompts. Use for text-to-video, image-to-video, multi-shot or dialogue scenes, reference-video transformation, video editing or continuation, product/logo/jewelry/fashion/social/UGC films, fight scenes, character stories, cooking tutorials, music videos, award or freeze effects, talking-character clips, drone or one-shot footage, long-form-to-short clipping plans, storyboards, prompt audits, camera/light/sound direction, race/chase/kinetic montage, or improving a video-generation node.

02 资源文件 30

03 SKILL.md 正文

Direct AI video creation

Use $operate-content-workflow for all inspection and mutations. This skill supplies creative decisions; runtime capabilities remain the source of truth for what can be executed.

Route the task

  1. Inspect the project and capabilities.
  2. If no concept, story, or scene exists, invoke $develop-creative-concept first.
  3. Classify the request as treatment/script, storyboard/animatic, text-to-video, image-to-video, reference transformation, edit/continuation, single shot, or multi-shot sequence.
  4. Build the brief and shot plan before adding generation nodes.
  5. Generate a low-cost proof shot when direction is uncertain; expand only after the direction is accepted.
  6. Review the actual output with $review-content-quality before recommending a retry.

Read only the references needed for the task:

Brief requirements

Record purpose, audience, publishing surface, duration, ratio, subject, environment, style, continuity anchors, action, camera, pacing, audio intent, references, constraints, deliverables, and acceptance criteria. Infer reversible defaults; ask only when a missing choice changes cost, supplied references, or the core deliverable.

Keep each generator focused on one visually coherent shot unless the inspected mode explicitly supports multiple timestamped stages or shots in one generation. In that mode, put consecutive time windows inside the video prompt. Otherwise create independent shot nodes and return their intended order. Name the changed variable in every intentional variation.

Choose a generation mode

  • Select the method from the authorities the shot must preserve, using video-strategy-selection.md before considering model availability.
  • Use the strategy-specific prompt plan kind required by that method, even when strategy="automatic".
  • Use first/last-frame behavior only when the inspected mode exposes distinct frame roles and both static endpoint plans are feasible.
  • Use omni-reference or video transformation only when the inspected input slots accept the required media kinds.
  • Prefer a short low-cost validation shot before a larger set when creative direction is uncertain.

When materializing the plan on the canvas, use $operate-content-workflow's high-level configure_video_shot contract. Pass concrete asset IDs for asset collections, reference upstream image/video generators by future output, and let capability resolution compile business roles to provider slots. Use low-level configure_generation only to preserve or repair an existing technical role configuration. Never express first frame, last frame, visual reference, motion reference, or source video as an unlabelled edge.

This method is an original Vetta adaptation informed by Generative-Media-Skills (MIT), visual-skills by Serge Shima (CC BY 4.0, https://github.com/smixs/visual-skills), and ViMax (MIT).

04 同仓库更多 Skills

openvetta
openvetta/open-vetta

publish-ability

Publish a skill, MCP server, plugin, or bundle to the Vetta ability marketplace. Use when the user asks to upload/publish/submit an ability, put a plugin on the marketplace, share a skill with other users, or migrate an existing package into Vetta.

★ 299
openvetta
openvetta/open-vetta

install-ability

在本机 Vetta 中安装能力——插件(plugin)、Skill 与 MCP 服务器。用户说「装一个插件 / install plugin / 从能力市场安装 / 加一个 MCP server / 配置 MCP / 装个 skill」,或插件装完没生效、需要 reload、pending 版本没应用时使用。不用于安装 npm 依赖、IDE 插件、浏览器扩展,也不用于开发插件(那是 plugin-workbench)或创作 Skill(那是 create-skill)。

★ 299
openvetta
openvetta/open-vetta

create-skill

Create or update Vetta-compatible Agent Skills. Use when the user asks to create a skill, add a SKILL.md workflow, register a skill in skills-manifest.json, package a skill in a Vetta plugin, or add a built-in Vetta skill preset.

★ 299
openvetta
openvetta/open-vetta

vetta-ui-design

Build and edit design documents (.vetd) on the Vetta design canvas — app screens, landing pages, slides, posters, infographics. Use when the user asks for a UI design, mockup, screen, deck, or poster, or attaches a design frame/element from the canvas. Frames are real React (TSX) routes, not pictures. Do NOT use for ordinary front-end work in the user's own codebase — building, styling or reworking pages of a real app is that repo's job, and none of the vetd_* tools or canvas rules here apply to it.

★ 299
openvetta
openvetta/open-vetta

remotion-video

Create, modify, and render Remotion React video projects in the current Vetta conversation workspace. Use for requests to make programmatic videos, motion graphics, animated explainers, social videos, data videos, or to edit an existing Remotion composition and export it as MP4.

★ 299
openvetta
openvetta/open-vetta

plugin-workbench

Create, implement, build, pack, install, reload, and manage Vetta desktop plugins for non-developers. Use whenever the user wants a Vetta plugin, plugin scaffolding, apply a `.vettapkg` package to Vetta, edit plugin.json name/guidingWords, or debug plugin load/install. Requires the Plugin Workbench input-bar toggle (hard isolation). The plugin handbook ships inside the project's own @vetta-org/plugin-sdk; locate it with the bundled CLI's docs command.

★ 299