WorkRally Skills overview

What this Skill package is

WorkRally CLI is an Agent-oriented production toolchain for animated stories and video creation. It connects projects, series, shots, uploads, assets, canvas layout, AI image generation and AI video generation.

The workflow starts with story and character assets, then moves through keyframes, animation, asset management and canvas composition instead of stopping at a single image or video API call.

Core capabilities

AI images: Kontext, multi-reference input and canvas placeholders
AI video: text, first/last frame, subject-to-video, polling and batches
Projects: project, series, shot CRUD, character recognition and model config
Assets: upload -> asset create -> material add
Canvas: Yjs collaboration, eight node types and incremental build-draft updates

Installation and access

npm install -g workrally
workrally auth login
workrally auth status

After authentication, use the workrally CLI as the stable entry point instead of manually composing many backend requests.

Four concepts to understand first

• Project: the top-level container for assets. Assets must belong to a project to appear in the web interface.
• Canvas: the infinite layout space for nodes, boards, text, media and sketches.
• Asset: the project-level file record created after upload.
• Material: the tree view that organizes characters, props, scenes and folders.

Tutorial map

• AI generation: images, video, placeholders, polling and model discovery.
• Uploads and assets: upload, asset creation and material organization.
• Series and shot workflow: series, shots, character recognition and batch generation.
• Infinite canvas: node types, boards, incremental merge and overwrite modes.
• Common pitfalls: project/canvas confusion, expired URLs, hard-coded models and invalid nodes.

Choose the right tutorial

• Generating images or videos: start with AI generation.
• Uploading files or building an asset tree: read uploads and assets.
• Building episodes, shots or batch scenes: read the series workflow.
• Arranging media on a board: read the infinite canvas guide.
• Seeing an error or unexpected result: start with common pitfalls.