canvas-agent-skill

What problem does it solve?

Manually creating, editing, versioning, annotating, and recovering tldraw Canvas documents is time-consuming and error-prone, especially when integrating with AI workflows for visual collaboration. This Skill eliminates that manual overhead by enabling AI agents to handle all Canvas document operations reliably.

Core Features & Use Cases

• Full Canvas Lifecycle Management: Create, inspect, patch, version, review, annotate, archive, and recover Canvas documents via agent-callable middleware tools, with guardrails to avoid invalid operations like inventing document IDs or unintended deletions.
• AI Image Integration: Generate images via Seedream AIGC and insert them directly into Canvas AI image holders, or add user-provided images to documents for design mockups, mood boards, or visual feedback.
• Use Case: Product and design teams can use this Skill to automatically generate annotated design iterations in Canvas, save versioned snapshots for stakeholder review, and restore previous versions if feedback requires reverting changes.

Quick Start

Ask your AI agent to create a new Canvas document titled "Q3 Product Roadmap", add a generated market trend image to the first page, and save a versioned snapshot for team review.