creating-windsurf-packages
About
This skill helps developers create Windsurf rules by providing the required format specifications. It enforces plain markdown without frontmatter, a 12,000 character limit, and single-file structure in `.windsurf/rules`. Use it when preparing custom rules to ensure compatibility with Windsurf's constraints.
Quick Install
Claude Code
Recommended/plugin add https://github.com/pr-pm/prpmgit clone https://github.com/pr-pm/prpm.git ~/.claude/skills/creating-windsurf-packagesCopy and paste this command in Claude Code to install this skill
Documentation
Creating Windsurf Packages
Overview
Windsurf uses a single .windsurf/rules file containing plain markdown instructions with NO frontmatter.
CRITICAL CONSTRAINTS:
- No frontmatter - Pure markdown only
- 12,000 character limit - Hard limit enforced by Windsurf
- Single file - All rules in one
.windsurf/rulesfile
Quick Reference
| Aspect | Requirement |
|---|---|
| Format | Plain markdown |
| Frontmatter | None (forbidden) |
| Character limit | 12,000 max |
| File location | .windsurf/rules (single file) |
Creating Rules
Plain markdown with optional H1 title and organized sections:
# React Development Guidelines
Guidelines for building React applications in this project.
## Component Structure
- Use functional components with hooks
- Keep components under 200 lines
- Extract logic into custom hooks when appropriate
- Co-locate styles with components
## State Management
We use Zustand for global state:
- Create stores in `src/stores/`
- Use selectors to prevent unnecessary re-renders
- Keep stores focused on single concerns
\`\`\`typescript
// Good: Focused store
const useAuthStore = create((set) => ({
user: null,
login: (user) => set({ user }),
logout: () => set({ user: null }),
}));
\`\`\`
## Testing
- Write tests alongside components (`.test.tsx`)
- Use React Testing Library
- Test user behavior, not implementation details
- Aim for 80% coverage on new code
Character Budget Tips
To stay under 12,000 characters:
- Focus on project-specific patterns - AI already knows general best practices
- Use concise language - Every word counts
- Limit code examples - Only essential patterns
- Skip obvious practices - Don't repeat what AI knows
- Reference external docs - Link instead of repeating
Example: Project-Specific Context
# Project Architecture
## Tech Stack
- **Frontend**: React 18 + TypeScript + Vite
- **Styling**: Tailwind CSS
- **State**: Zustand
- **Routing**: React Router v6
- **API**: REST with axios
## Directory Structure
\`\`\`
src/
components/ # Reusable UI components
features/ # Feature-specific code
hooks/ # Custom React hooks
stores/ # Zustand stores
utils/ # Helper functions
types/ # TypeScript types
\`\`\`
## Coding Conventions
- Use PascalCase for components
- Use camelCase for functions/variables
- Use kebab-case for file names
- Export components as named exports
## API Integration
All API calls go through `src/api/client.ts`:
\`\`\`typescript
import { apiClient } from '@/api/client';
// Use the client
const users = await apiClient.get('/users');
\`\`\`
## Environment Variables
Access via `import.meta.env`:
- `VITE_API_URL` - Backend API URL
- `VITE_APP_ENV` - Environment (dev/staging/prod)
Content Format
Standard markdown including:
- H1 title: Main heading (optional)
- H2/H3 sections: Organize content
- Lists: Unordered and ordered
- Code blocks: With language specifiers
- Standard markdown: Bold, italic, links
Common Mistakes
| Mistake | Fix |
|---|---|
| Adding YAML frontmatter | No frontmatter allowed - plain markdown only |
| Exceeding 12,000 chars | Prioritize project-specific content, trim aggressively |
| Multiple files | All content in single .windsurf/rules file |
| Generic best practices | Focus on project-specific patterns |
| Verbose examples | Keep examples minimal and focused |
Checking Character Count
Use this command to check character count:
wc -m .windsurf/rules
Stay well under 12,000 to leave room for updates.
What to Include
High Priority (include these):
- Project-specific tech stack
- Directory structure and naming conventions
- API patterns unique to your project
- Custom tooling and scripts
- Environment-specific configuration
- Team conventions and workflows
Low Priority (skip these):
- General programming best practices
- Language syntax explanations
- Framework basics (React, TypeScript)
- Obvious code quality rules
- Verbose explanations of standard patterns
Example: Full Stack Project
# TaskMaster Development Guide
## Architecture
### Frontend
- React 18 with TypeScript
- Vite for build tooling
- Zustand for state management
- React Query for server state
- Tailwind CSS for styling
### Backend
- Node.js with Express
- PostgreSQL with Prisma ORM
- WebSocket for real-time features
- Redis for caching and pub/sub
- JWT for authentication
## File Structure
\`\`\`
src/
components/ # Reusable UI components
features/ # Feature-based modules
hooks/ # Custom React hooks
lib/ # Utility functions
pages/ # Route pages
types/ # TypeScript types
\`\`\`
## Development Workflow
1. Create feature branch from `main`
2. Write tests first (TDD)
3. Implement feature
4. Run `pnpm test` and `pnpm lint`
5. Create PR with description
6. Merge after approval
## Testing
- Use Vitest for unit tests
- Use Playwright for E2E tests
- Aim for 80% coverage on new code
- Mock external dependencies
Compression Techniques
Verbose (100+ words):
When you are working with React components, it's very important to remember
that you should always use functional components with hooks instead of class
components. This is because hooks provide a more modern and flexible way to
manage state and side effects. Additionally, you should keep your components
small and focused on a single responsibility...
Concise (30 words):
## React Components
- Use functional components with hooks (no classes)
- Keep under 200 lines
- Single responsibility
- Custom hooks for complex logic
Migration from Other Formats
When converting to Windsurf:
- Strip all frontmatter - Remove YAML headers completely
- Combine multiple files - Merge into single document
- Prioritize content - Keep project-specific, remove generic
- Trim examples - Only essential code samples
- Monitor length - Check character count regularly
Validation
Documentation: /Users/khaliqgant/Projects/prpm/app/packages/converters/docs/windsurf.md
Schema location: /Users/khaliqgant/Projects/prpm/app/packages/converters/schemas/windsurf.schema.json
Best Practices
- Be concise - 12,000 character limit means prioritize
- Single file - Combine all project context cohesively
- Clear structure - Use headers for scannable sections
- Real examples - Show actual code patterns from project
- Update regularly - Keep in sync with architecture changes
- No frontmatter - Plain markdown only, no YAML
Remember: Windsurf uses plain markdown with NO frontmatter. 12,000 character limit. Single .windsurf/rules file.
GitHub Repository
Related Skills
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
evaluating-llms-harness
TestingThis Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.
llamaguard
OtherLlamaGuard is Meta's 7-8B parameter model for moderating LLM inputs and outputs across six safety categories like violence and hate speech. It offers 94-95% accuracy and can be deployed using vLLM, Hugging Face, or Amazon SageMaker. Use this skill to easily integrate content filtering and safety guardrails into your AI applications.
langchain
MetaLangChain is a framework for building LLM applications using agents, chains, and RAG pipelines. It supports multiple LLM providers, offers 500+ integrations, and includes features like tool calling and memory management. Use it for rapid prototyping and deploying production systems like chatbots, autonomous agents, and question-answering services.
