arch-viewlisted

# Architecture View Generate cross-repository architectural views by aggregating `catalog-info.yaml` metadata from all repositories using parallel subagents. ## Purpose Create visual architectural representations (Mermaid diagrams, markdown tables) that show how services fit together, how requests flow through gateways, how events propagate, and how services group by domain/team. ## When to Use Trigger this skill when: - User asks to "show service dependency graph" or "map the architecture" - User wants to understand "how services connect" - User asks about "request flows" or "event topology" - User wants to "group services by domain/team" - User mentions "cross-repo architecture" or "system architecture" ## Workflow ### Phase 1: Discover Repositories Find all repositories to analyze: 1. **Check `repos/` directory** - Local cloned repositories 2. **Optionally fetch from GitHub** - Use `gh repo list Astrabit-CPT` for full list 3. **Filter for active repos** - Skip archived or template repos ### Phase 2: Parallel Metadata Collection Use subagents IN PARALLEL to read each repository's `catalog-info.yaml`: ``` For each repo: Launch subagent with: "Read catalog-info.yaml from [repo_path] and return the parsed content" ``` **Parallel processing strategy:** - Launch 5-10 subagents simultaneously - Collect all results - Handle missing metadata gracefully (skip or note as missing) ### Phase 3: Aggregate and Build Model Combine all metadata into a unified model: ```py