docs-init

/docs:init — Initial Documentation Scan

Usage

/docs:init

Behavior

Performs a one-time initial documentation scan of the entire codebase and generates a documentation skeleton. Run once at project setup, then use /docs:sync for incremental updates.

Execution Steps

1. Scan entire codebase structure. Walk the project directory tree and catalog:

   • All source files, grouped by module/service/component
   • Entry points (main files, route definitions, controllers)
   • Configuration files
   • Existing documentation (if any)

2. Identify modules, services, components, routes. Build an inventory:

   • Backend services and their responsibilities
   • Frontend components and their hierarchy
   • API routes/endpoints with HTTP methods
   • Database models/schemas
   • Shared utilities and libraries
   • Third-party integrations

3. Generate documentation skeleton with the following sections:

   • Project Overview: what the project does, tech stack, high-level architecture
   • Module Index: list of all modules/services with one-line descriptions
   • API Reference Stubs: endpoint list with method, path, and placeholder for description
   • Setup Guide: detected prerequisites, environment variables needed, build/run commands
   • Component Index (if frontend): component tree with props/events stubs

4. Write to docs/ directory. Create:

   docs/
   ├── overview.md
   ├── modules.md
   ├── api-reference.md
   ├── setup.md
   └── components.md  (if frontend detected)
   

When to Use

• First time setting up the toolkit on an existing project
• When joining a new codebase and need to understand the structure
• After a major restructuring that invalidated previous docs

Output

Documentation skeleton in docs/ directory. Each file contains accurate structure with placeholder content where details need to be filled in. Use /docs:sync going forward to keep docs up to date with code changes.