code-doc

What problem does it solve?

This Skill eliminates the manual burden of creating and maintaining codebase documentation, providing both human-readable Markdown and machine-readable JSON output across multiple programming languages. It helps you quickly understand unfamiliar codebases and generate optimized context for AI assistants.

Core Features & Use Cases

• Multi-Language Support: Documents Python, JavaScript/TypeScript, Go, HTML, CSS, and more, with AST parsing and complexity metrics.
• AI-Enhanced Documentation: Uses multi-agent AI consultation (e.g., Cursor-Agent, Gemini) to synthesize architecture and AI assistant context.
• Code Metrics & Analysis: Includes AST parsing, dependency analysis, and complexity metrics to highlight code quality and refactoring needs.
• Use Case: Onboard a new developer to a legacy project by generating comprehensive ARCHITECTURE.md and AI_CONTEXT.md files, giving them an instant, AI-synthesized overview of the system's design and key components.

Quick Start

Option 1 (Recommended): AI-Enhanced Documentation

Gathers research from AI tools, main agent synthesizes ARCHITECTURE.md and AI_CONTEXT.md

sdd doc analyze-with-ai <directory> --name ProjectName --version X.Y.Z --verbose

Option 2: Structural Documentation Only (no AI required)

Generates DOCUMENTATION.md and documentation.json

sdd doc generate <directory> --name ProjectName --version X.Y.Z --format both --verbose

Validate the generated JSON documentation

sdd doc validate-json ./docs/documentation.json