A tool designed to transform vague video concepts into precise, actionable scripts. It targets creators struggling with pre-production, acting as an AI director to elicit detailed shot breakdowns, pacing, and content specifications. The primary benefit is streamlining the ideation-to-scripting phase, producing a video-spec.md file ready for rendering by tools like HyperFrames.

How It Works

This project integrates with AI coding environments (Codex, Claude Code, Cursor) to facilitate a conversational script-building process. The video-spec-builder skill prompts users with targeted questions, akin to a human director, to clarify video objectives, audience, core messages, and visual elements. It iteratively refines the user's input, pushing for concrete details until a comprehensive video-spec.md file is generated. This script details each shot, its duration, content, and transitions, serving as direct input for downstream rendering tools.

Quick Start & Requirements

• Primary install: npx skills add feicaiclub/video-spec-builder (use -g for global installation).
• Prerequisites: Node.js 18 or newer. Requires the HyperFrames renderer to be installed (npx skills add heygen-com/hyperframes).
• Environment: Best used within AI-powered coding environments like Codex, Claude Code, or Cursor.
• Links: Official documentation is implicitly accessed via the AI environments.

Highlighted Details

• Generates precise, second-timed video-spec.md shot-by-shot scripts from conversational input.
• Employs an AI-driven directorial approach to extract detailed requirements and overcome vague ideas.
• Includes a bundled "Spec Mono" custom theme for a specific dark, geometric, engineered aesthetic.
• Designed for seamless integration with the HyperFrames video rendering engine.

Maintenance & Community

No specific details regarding maintainers, community channels (e.g., Discord, Slack), or roadmap were found in the provided README excerpt.

Licensing & Compatibility

• License: MIT.
• Compatibility: Permissive MIT license allows for commercial use and integration into closed-source projects. Compatible with Node.js environments and specified AI coding assistants.

Limitations & Caveats

The tool focuses solely on script generation; it does not create video assets like illustrations, live-action footage, or photorealistic images. The downstream HyperFrames renderer is limited to elements constructible via HTML, CSS, and code, excluding complex artwork or real-world captures. AI-generated voiceovers may exhibit a noticeable machine quality, and users must supply all necessary media assets (clips, images, music) for rendering.