tldraw demos Claude for docs
tldraw demoed Anthropic’s Claude generating a release‑process diagram from GitHub workflows, showcasing how multimodal AI can turn CI configs into visual engineering docs. It’s a concrete example of using models to automate documentation and reduce manual handoffs in release planning. (x.com)
A community-maintained repo named tldraw-mcp wires tldraw to Anthropic’s Claude via the Model Context Protocol so Claude can create and edit canvas shapes programmatically. (github.com)) The tldraw organization’s codebase includes a GitHub Actions workflow file called claude.yml, indicating an integration point that can run Claude-driven automation inside their CI. (github.com)) Inside the tldraw repo there’s a.claude/skills/write-release-notes directory that contains a Claude skill for generating MDX release logs from merged PRs and git history. (github.com)) Independent projects such as MOZARTINOS/claude-code-skill-tldraw-diagram produce VS Code–compatible.tldr files from JSON, enabling programmatic export of AI-generated canvas diagrams. (github.com)) Marketplace listings and skill directories (MCPMarket / SkillsMP) host “release notes writer” and related Claude skills that parse PRs and commit ranges to auto-generate release documentation. (mcpmarket.com)) Anthropic’s Claude Code documentation describes GitHub Actions integrations where a simple @claude trigger can run code analysis and automated edits, the same mechanism tldraw projects use to convert CI config into visual artifacts. (code.claude.com))
