Democratizing AI capabilities through open-source infrastructure.

We build production-ready tools that empower developers to create the next generation of AI-powered applications. From document processing to knowledge retrieval, we're making advanced AI accessible to everyone.

• 🤖 300+ stars on flagship PDF Reader MCP server
• 📦 8,000+ combined downloads across npm packages
• 🎯 Featured on mcpservers.org, lobehub.com, pulsemcp.com
• ✅ 94%+ test coverage on production systems
• 🔐 Security validated by mseep.ai
• 🚀 5-10x performance improvements in document processing

Production-grade PDF processing server for AI agents using Model Context Protocol.

• Performance: Processes 50-page PDFs in seconds with parallel batch processing
• Features: Text & image extraction, smart content ordering, metadata access
• Speed: 5-10x faster than sequential processing
• Integration: Works with Claude Desktop, Cline, and all MCP-compatible tools
• Security: Validated implementation with comprehensive test suite

npx @sylphx/pdf-reader-mcp

Secure, efficient, and token-saving filesystem access for AI agents.

• Token Efficiency: Batch operations reducing AI-server round trips
• Security: Root directory confinement with permission controls
• Features: Multi-file operations, recursive processing, surgical editing
• Distribution: Available on npm, Docker Hub, and MCP directories

npx @sylphx/filesystem-mcp

📝 Read our Medium article

Retrieval-Augmented Generation server with local vector database and embeddings.

• Privacy: Local ChromaDB storage with Ollama embeddings
• Performance: Fast semantic search across project documentation
• Integration: Built with Google Genkit framework
• Features: Document indexing, query, removal, and listing

Enable AI agents to access project-specific knowledge while maintaining full data control.

Type-safe, annotation-based Firestore ODM for Dart/Flutter with code generation.

• Architecture: Clean architecture with zero runtime reflection
• Performance: Lightning-fast code generation using callables
• Features: Full type safety, pagination, streaming, bulk operations
• Documentation: Comprehensive docs site at SylphxAI.github.io/firestore_odm

MEP (Minimal Effective Prompt) - A paradigm shift in AI-assisted development.

• Innovation: ~90% reduction in prompt composition time
• Technology: StarCoder2 tokenization for 70+ languages
• Features: Automatic context gathering, semantic codebase search
• Value: Faster onboarding, consistent quality, multi-language support

Stop spending time crafting prompts. Let Flow gather context automatically.

TypeScript • Node.js • Dart • Bun • Docker

Model Context Protocol (MCP) • ChromaDB • Ollama
Vector Databases • RAG Systems • LLM Integration

ESLint Flat Configs • Strict TypeScript • Prettier
Code Generation • Build Automation • CI/CD

React • Next.js • Flutter • Vue.js • Google Genkit
Firebase • Firestore • GitHub Actions

Recognized the strategic importance of Model Context Protocol immediately after Anthropic's announcement (Nov 2024) and became a major contributor within months.

• ✅ 94%+ test coverage on flagship projects
• 🔐 Security validation and best practices
• 📚 Comprehensive documentation for every project
• 🤖 Automated CI/CD pipelines

• 5-10x faster PDF processing
• Token-efficient batch operations
• Smart caching and optimization
• ~12,933 ops/sec throughput

• MIT licensing for community benefit
• Active issue tracking and support
• Knowledge sharing via Medium articles
• Contributing to MCP ecosystem growth

• Type-safe code generation
• Zero-dependency utilities
• Clear, actionable documentation
• Easy integration (npx/bunx)

• @sylphx/pdf-reader-mcp - PDF processing MCP server
• @sylphx/filesystem-mcp - Filesystem access MCP server
• @sylphx/mcp-rag-server - RAG server with ChromaDB + Ollama
• @sylphx/media-curator - Intelligent media organizer with LSH deduplication
• @sylphx/craft - Fast immutable state library (1.4-35x faster than immer)
• @sylphx/zen - Minimal state management (1.45 KB, 1.7-45x faster)
• @sylphx/silk - Zero-runtime CSS framework (92% smaller than Panda)
• @sylphx/ai-sdk-provider-claude-code - Claude Code provider for Vercel AI SDK
• @sylphx/mcp-tools - Modular MCP tools toolkit

• firestore_odm - Type-safe Firestore ODM
• firestore_odm_annotation - ODM annotations
• firestore_odm_builder - Code generator

• sylphx/filesystem-mcp - Containerized filesystem server
• sylphx/pdf-reader-mcp - PDF processing server
• More coming soon...

Revolutionary approach to AI-assisted development:

• 🎯 Focus: Minimal input, maximum context
• 🤖 Automation: Semantic codebase analysis
• 🧠 Intelligence: Pattern recognition across languages
• ⚡ Efficiency: 90% reduction in prompt engineering time

Catch errors at compile-time, not runtime:

• 🔒 Safety: Full type inference and checking
• 🚀 Speed: Reduced boilerplate, faster development
• 🌐 Multi-Platform: TypeScript, Dart, C# support
• 💡 DX: Excellent developer experience

Your data stays yours:

• 🔐 Local Processing: ChromaDB + Ollama
• 🏠 On-Premises: No cloud dependencies
• 🛡️ Security: Sandboxing and permission controls
• 📊 Transparency: Open-source, auditable code

• 🌟 Star our repos - Help others discover our tools
• 🐛 Report issues - Help us improve quality
• 💡 Suggest features - Shape the future of MCP
• 🤝 Contribute - Pull requests welcome!

• 📝 Medium - Technical deep-dives and announcements
• 🐙 GitHub - Watch releases and discussions
• 📦 npm - Follow @sylphx for package updates

• 📚 Firestore ODM Docs
• 📝 Medium Articles
• 🔧 Individual project READMEs

• Stop Fighting Firestore Schemas
• Empower Your AI Agent with Filesystem MCP

Location: London, United Kingdom Founded: 2025 Status: UK Registered Company

We believe AI agents will revolutionize software development. By building robust, secure, and efficient infrastructure, we're enabling developers worldwide to harness AI's full potential while maintaining control over their data and workflows.

🎯 Production-Ready Quality → Ship code you're proud of
🌍 Open Source First → Community over competition
⚡ Performance Matters → Every millisecond counts
🔐 Security by Design → Trust through transparency
💡 Developer Experience → Tools that developers love
🤝 Knowledge Sharing → Learn together, grow together

# Via npx (easiest)
npx @sylphx/pdf-reader-mcp

# Via Docker
docker run -v /path/to/pdfs:/pdfs sylphx/pdf-reader-mcp

# Via npx
npx @sylphx/filesystem-mcp

# Via Bun
bunx @sylphx/filesystem-mcp

# Clone and run
git clone https://github.com/SylphxAI/rag-server-mcp
cd rag-server-mcp
npm install
npm start

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "pdf-reader": {
      "command": "npx",
      "args": ["-y", "@sylphx/pdf-reader-mcp"]
    },
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@sylphx/filesystem-mcp"]
    }
  }
}

• 📄 Enhance PDF Reader with more extraction features
• 🗂️ Filesystem MCP v2 with improved performance
• 🧠 RAG Server optimization and new embedding models
• 📚 Comprehensive documentation expansion

• 🔗 More MCP servers (Slack, Jira, Linear integrations)
• 🎓 Developer tutorials and workshops
• 💼 Enterprise offerings and support
• 🌐 Community Discord server

• 🏢 Complete AI development platform
• 🎓 MCP certification programs
• 🤝 Enterprise consulting services
• 🌍 Global developer community