| allowed-tools | Read, Glob, Grep, Write, MultiEdit, TodoWrite, Bash |
| description | Generate comprehensive technical stack documentation from codebase analysis |
| name | create-tech-stack |
Generate Tech Stack Documentation
User Request: $ARGUMENTS
Context
- Project root: !
pwd
- Package.json: @package.json
- Node version: !
node --version 2>/dev/null || echo "Node.js not found"
- TypeScript config: @tsconfig.json
- Database schema: !
ls -la prisma/schema.prisma 2>/dev/null || echo "No Prisma schema found"
- Existing docs: !
ls -la .claude/rules/*.md 2>/dev/null || echo "No docs yet"
Goal
Create comprehensive Tech Stack Documentation based on deep codebase analysis. Document all technologies, frameworks, libraries, development tools, deployment strategies, and implementation patterns with specific versions and configurations.
Process
1. Automated Technical Discovery
- Parse package.json for all dependencies
- Analyze configuration files (tsconfig, vite.config, next.config, etc.)
- Detect database setup (Prisma, Drizzle, TypeORM, etc.)
- Identify testing frameworks and tools
- Scan for CI/CD configurations
- Check deployment configurations
2. Deep Code Analysis
Examine codebase for:
- Architecture Patterns: Monorepo structure, module organization
- Framework Usage: Next.js app router vs pages, API routes
- State Management: Zustand, Redux, Context API patterns
- Styling Approach: Tailwind, CSS modules, styled-components
- Type Safety: TypeScript strictness, validation libraries
- API Design: REST, GraphQL, tRPC implementation
- Authentication: Auth libraries and session management
- Testing Strategy: Unit, integration, E2E test patterns
3. Interactive Technical Q&A
Ask 4-6 deployment and infrastructure questions:
- Use numbered/lettered options
- Focus on non-discoverable information
- Gather hosting, monitoring, and workflow details
4. Generate Comprehensive Documentation
Create detailed tech stack document with:
- Specific version numbers
- Configuration examples
- Command references
- Architecture diagrams (when applicable)
5. Save and Organize
- Create
.claude/rules/ if needed
- Save as
2-tech-stack.mdc
Technical Questions Template
🚀 Deployment & Infrastructure
1. Where is your application currently deployed?
a) Vercel - Next.js optimized hosting
b) AWS - EC2, Lambda, or containerized
c) Railway/Render - Modern PaaS providers
d) Self-hosted - VPS or on-premise
e) Other - Please specify
f) Not deployed yet - Still in development
2. How is your database hosted?
a) Managed service (Supabase, PlanetScale, Neon, etc.)
b) Cloud provider (AWS RDS, Google Cloud SQL, etc.)
c) Self-hosted (Docker, VPS, etc.)
d) Local only - No production database yet
📊 Monitoring & Operations
3. What observability tools do you use?
a) Error tracking: Sentry, Rollbar, Bugsnag
b) Analytics: Vercel Analytics, Google Analytics, Plausible
c) Monitoring: Datadog, New Relic, custom solution
d) Logging: CloudWatch, LogTail, custom logs
e) None yet - Planning to add later
👥 Development Workflow
4. What's your Git workflow?
a) Feature branches with PR reviews
b) Trunk-based development
c) GitFlow with release branches
d) Direct to main (solo project)
5. How do you manage environments?
a) Multiple deployments (dev, staging, prod)
b) Preview deployments for PRs
c) Single production environment
d) Local development only
🔐 Additional Services
6. Which external services do you integrate with?
Document Structure
The generated document must follow this technical structure:
Overview
- Brief description of the application's technical nature
- Technology stack summary
- Architecture approach (monolith, microservices, etc.)
Programming Language & Runtime
- Primary programming language and version
- Runtime environment and version
- Type system and language features used
Frontend
- UI Framework/Library and version
- Styling approach and frameworks
- Component libraries and design systems
- State management solutions
- Build tools and bundlers
- Browser support and compatibility
Backend
- Backend framework and architecture
- API design (REST, GraphQL, tRPC, etc.)
- Authentication and authorization
- Middleware and security
- File handling and uploads
Database & Storage
- Database type and version
- ORM/Query builder
- Schema management and migrations
- Caching solutions
- File storage solutions
- Data backup and recovery
Development Tools & Workflow
- Package manager
- Code formatting and linting
- Type checking and compilation
- Testing frameworks and strategies
- Development server and hot reload
- Version control workflow
Deployment & Infrastructure
- Hosting platform and services
- Build and deployment pipeline
- Environment configuration
- Domain and DNS management
- SSL/TLS and security
- Monitoring and logging
External Integrations
- Third-party APIs and services
- Payment processing
- Email services
- Analytics and tracking
- Error monitoring
- Performance monitoring
Quality Assurance & Testing
- Testing strategy and frameworks
- Code coverage tools
- End-to-end testing
- Performance testing
- Security testing
- Code review process
Schemas & Data Models
- Database schema (if applicable)
- API schemas and validation
- Type definitions and interfaces
- Data relationships and constraints
Target Audience
The document should serve:
- Developers joining the project who need technical onboarding
- DevOps engineers setting up infrastructure and deployment
- Technical architects evaluating or improving the tech stack
- Security teams understanding the technical landscape
- Future maintainers who need to understand technical decisions
The language should be technical, precise, and include specific version numbers and configuration details.
Documentation Principles
DO Include:
- Exact Versions: Lock file versions, not just ranges
- Configuration Examples: Actual config snippets from the project
- Command Reference: All npm scripts and their purposes
- Setup Instructions: Step-by-step for new developers
- Architecture Decisions: Why specific technologies were chosen
- Integration Details: How services connect and communicate
DON'T Include:
- Generic Descriptions: Avoid Wikipedia-style explanations
- Outdated Information: Only document what's actually used
- Wishful Thinking: Document current state, not future plans
- Sensitive Data: No API keys, secrets, or credentials
- Redundant Info: Link to official docs instead of copying
Output
- Format: Markdown (
.mdc)
- Location:
.claude/rules/
- Filename:
2-tech-stack.mdc
Execution Steps
1. Automated Analysis Phase
- Read package.json and lock files
- Scan for configuration files
- Detect framework patterns
- Identify database setup
- Find test configurations
2. Manual Discovery Phase
- Read key source files to understand architecture
- Check for API route patterns
- Analyze authentication implementation
- Review deployment configurations
3. Interactive Q&A
- Present deployment and infrastructure questions
- Use checkboxes for multi-select options
- Wait for user responses
4. Document Generation
- Start with discovered information
- Incorporate user responses
- Add specific configuration examples
- Include all npm scripts with descriptions
5. Save and Update
mkdir -p .claude/docs
6. Next Steps
- Recommend: "Should I create an app design document to complement this technical documentation?"
Example Usage
/project:create-tech-stack
/project:create-tech-stack Focus on deployment and CI/CD setup
Sample Output Structure
# Tech Stack Documentation
## Overview
- **Framework:** Next.js 14.2.5 (App Router)
- **Language:** TypeScript 5.5.3
- **Database:** PostgreSQL with Prisma ORM
- **Deployment:** Vercel with preview deployments
## Commands Reference
### Development
- `pnpm dev` - Start Next.js dev server on port 3000
- `pnpm build` - Build production bundle
- `pnpm typecheck` - Run tsc --noEmit
### Database
- `pnpm db:generate` - Generate Prisma client
- `pnpm db:push` - Push schema changes to database
# ... continue with full documentation
