Agent skill that turns natural language, code, Terraform/K8s, SQL, OpenAPI, AsyncAPI, Protobuf and GraphQL sources into editable, tested draw.io architecture diagrams: incremental sync, multi-view projection, drift diff, CI architecture tests, whiteboard derasterize, interactive HTML/PPTX/Mermaid exports.
Python agent-skillsarchitecturearchitecture-diagramasyncapibpmnc4-modelclaude-codediagramdiagrams-as-codedrawioflowchartgithub-actionsgraphqlimage-to-diagramkubernetesmcp-serverprotobufsysmlterraformuml
docs(changelog): note the rendered-output guidance under Unreleased
Agents365-aia7e2d00bc180
diff --git a/CHANGELOG.md b/CHANGELOG.mdindex 210923e..90f4e4f 100644--- a/CHANGELOG.md+++ b/CHANGELOG.md@@ -18,6 +18,16 @@ semantic-ish versioning (the `metadata.version` field in`diagramctl.py build` (auto-detected from `.graphql` / `.gql`) and the MCPserver.+### Changed++- **Rendered-output pitfalls documented**: `references/xml-authoring.md` gains a+ decision-diamond section (outward vertex exits, or straight `edgeStyle=none`+ lines from the lower edge midpoints; full-precision `entryX`/`exitX` pins;+ `endArrow=block;endSize=8`; micro-labels only on the line), and+ `references/troubleshooting.md` gains a DOM-verification workflow (dump the+ viewer DOM and assert `<path d>` segments and label anchors) with two new+ failure-mode rows. Creation workflow step 5 now points at that fallback.+## [3.3.0] - 2026-09-12### Added