What problem does it solve? Understanding how a project is structured — its systems, containers, and components — requires reading scattered manifests, deploy descriptors, and documentation. This Skill automates that by deriving a C4 architecture model from repository evidence and rendering it as one interactive HTML page the reader can drill down through, from system context to containers to components. ## Core Features & Use Cases - Repo classification and evidence grounding: Classifies the repository as software, plugin/agent-asset, or documents, then harvests manifests, deploy descriptors, ADRs, and READMEs, attaching an evidence pointer to every modelled element. - Reconciled model with sticky decisions: Writes diagram/c4-model.json in a Structurizr-compatible schema, preserving hand-edited fields, recorded judgement calls, and grid positions across re-runs while retiring unevidenced elements instead of deleting them. - Template-rendered interactive HTML: Renders diagram/c4.html wholesale from a plugin template via token substitution, producing a self-contained page that works offline over file:// with no external dependencies. - Use Case: Point it at a microservices repo to produce a drill-down architecture page for onboarding, or run it in a docs-only repo to turn ADRs and design documents into a navigable C4 model. ## Quick Start Ask the assistant to build an interactive C4 architecture diagram for the current project, optionally specifying fresh or render mode.
