--- name: architecture-overview description: This skill should be used when the user asks for a high-level overview of the codebase, "explain the architecture", "what does this codebase look like", "how is this project structured", "what languages are used", or wants a quick mental model of a project's structure and size. --- # Architecture Overview Retrieve a high-level architectural overview of the codebase using the Constellation `code_intel` tool (from the `constellation` MCP server). ## How to Run Call the `code_intel` tool with this code: ```javascript const result = await api.getArchitectureOverview({ includeMetrics: true, includeModuleGraph: false }); return result; ``` ## Presenting Results 1. **Project Summary** - Primary language (`result.data.metadata.primaryLanguage`) and detected frameworks (`result.data.metadata.frameworks`) - Total files (`result.data.metadata.totalFiles`) and total symbols (`result.data.structure.symbols.total`) 2. **Language Distribution** - For each language in `result.data.metadata.languages`, show: - Language name (`language`) - File count (`fileCount`) - Percentage of codebase (`percentage`) 3. **Symbol Breakdown** - Top 10 symbol kinds from `result.data.structure.symbols.byKind` (functions, classes, interfaces, etc.) with counts - Exported symbol count (`result.data.structure.symbols.exported`) 4. **Dependency Hotspots** - Most connected files from `result.data.dependencies.internal.mostConnectedFiles` (path, incoming/outgoing counts); these are the hub modules of the codebase - Top external packages from `result.data.dependencies.external.topPackages` 5. **Quality Metrics** (only if `result.data.metrics` is present) - Average complexity and high-complexity count (`metrics.complexity`) - Maintainability score and issues (`metrics.maintainability`) - Test coverage percentage if available (`metrics.testCoverage`) 6. **Observations** (optional insights) - Note if it's heavily function-based vs class-based - Multi-language codebase characteristics - Concentration of connections in a few hub files Keep the output concise and scannable. Focus on giving a quick mental model of the codebase structure. **If error**, explain the error and provide guidance from the error response. The `constellation-troubleshooting` skill covers error codes in depth.