--- allowed-tools: [Read, Grep, Glob, Task, WebSearch, WebFetch] description: "Create clear, accurate technical documentation following project patterns" wave-enabled: true complexity-threshold: 0.5 --- # /hc:document - Technical Documentation ## Purpose Create clear, accurate technical documentation that follows existing project patterns and helps users succeed with the software. ## Usage ```bash /hc:document [target] [--type readme|api|guide|inline|reference] [--scope focused|comprehensive] [--framework nextra|docusaurus|vitepress|none] ``` ## Arguments - `target` - File, directory, function, or feature to document - `--type` - Documentation type: - `readme` - Project or module README files - `api` - API reference documentation - `guide` - User guides and tutorials - `inline` - Code comments and annotations - `reference` - Technical specifications and configurations - `--scope` - Documentation depth (focused for specific items, comprehensive for full coverage) - `--framework` - Documentation framework to use (optional, defaults to existing or none) ## Execution 1. **Pattern Analysis Phase** - Use @agent-tech-writer to analyze existing documentation patterns - Identify project's documentation style and conventions - Understand the codebase structure and implementation - Determine target audience and their needs 2. **Content Generation Phase** - @agent-tech-writer creates documentation following identified patterns - For API docs: Extract from actual implementation and annotations - For guides: Focus on real user tasks and common scenarios - For README: Follow project's existing structure or best practices - Ensure technical accuracy over comprehensive coverage 3. **Quality Assurance Phase** - Validate accuracy against actual code implementation - Check consistency with existing documentation - Ensure examples work and are practical - Verify accessibility and readability 4. **Integration Phase** - Place documentation in appropriate locations - Update cross-references and navigation - Ensure documentation fits naturally in project ## Claude Code Integration - **Primary Agent**: @agent-tech-writer for all documentation tasks - **Supporting Agents**: - @agent-architect: When needing system design context - @agent-coder: For extracting implementation details - @agent-designer: For UI component documentation - **MCP Servers**: - Tree-Sitter: Analyze code structure and extract signatures - Context7: Research best practices when patterns unclear - Memory: Store and retrieve documentation patterns - **Search Priority**: WebSearch > WebFetch for documentation standards - **Tools**: - Read/Grep: Analyze existing documentation - Task: Create, update, and manage documentation files - Task: Coordinate with other agents when needed ## Examples ```bash # Document a specific API endpoint /hc:document src/api/users.js --type api --scope focused # Create comprehensive project README /hc:document . --type readme --scope comprehensive # Generate user guide for a feature /hc:document src/features/auth --type guide # Add inline documentation to code /hc:document src/utils/validators.js --type inline # Build documentation site with Nextra /hc:document . --type guide --scope comprehensive --framework nextra ``` ## Best Practices - Always analyze existing patterns first - Focus on what users need to accomplish - Write from the user's perspective - Provide working, practical examples - Maintain technical accuracy - Let content drive structure, not framework features