juanre-ai-tools

AI Tools Plugin Marketplace for Claude Code

Python tools for AI development by Juan Reyero

Production-ready packages for databases, LLM integration, and document search. Each package includes multiple Claude Code skills providing guidance on best practices and patterns.

Quick Start

# Add marketplace
/plugin marketplace add juanre/ai-tools

# Install all skills for a package
/plugin install pgdbm@juanre-ai-tools
/plugin install llmemory@juanre-ai-tools
/plugin install llmring@juanre-ai-tools

# Or install individual skills
/plugin install pgdbm-shared-pool@juanre-ai-tools
/plugin install llmemory-hybrid-search@juanre-ai-tools
/plugin install llmring-streaming@juanre-ai-tools

What Are These Skills?

Claude Code skills are expert guidance systems that teach Claude how to work with your packages. When you ask Claude for help with databases, LLMs, or document search, these skills automatically activate and provide:

• ✅ Production-ready code examples
• ✅ Best practices and patterns
• ✅ Common pitfalls to avoid
• ✅ Architecture guidance
• ✅ Testing strategies

No manual needed - just ask Claude naturally and the right skills load automatically!

Available Packages

🗄️ pgdbm - PostgreSQL Database Management

Production-ready async PostgreSQL with migrations, connection pooling, and schema isolation.

Package on GitHub | PyPI

Skills included:

• pgdbm - Complete skill set (all 9 skills)
• pgdbm-choosing-pattern - Decision guide for which pattern to use
• pgdbm-shared-pool - Production pattern with connection pooling ⭐
• pgdbm-dual-mode - Create portable database libraries
• pgdbm-standalone - Standalone service pattern
• pgdbm-testing - Test your database code
• pgdbm-usage - Basic operations and queries
• pgdbm-core-api - Complete API reference
• pgdbm-migrations-api - Migrations API reference
• pgdbm-common-mistakes - Common pitfalls and how to avoid them

When to use: Building FastAPI apps, microservices, multi-tenant SaaS, or any production application needing PostgreSQL.

Key features:

• Async operations with asyncpg
• Automatic migrations with version control
• Connection pooling for production
• Schema isolation for multi-module apps
• Built-in testing utilities

# Install all pgdbm skills
/plugin install pgdbm@juanre-ai-tools

# Or just what you need
/plugin install pgdbm-shared-pool@juanre-ai-tools

🔍 llmemory - Document Memory with Vector Search

High-performance document memory system with hybrid search, multi-query expansion, and reranking.

Package on GitHub | PyPI

Skills included:

• llmemory - Complete skill set (all 5 skills)
• llmemory-basic-usage - Installation and basic operations
• llmemory-hybrid-search - Combining vector + BM25 search ⭐
• llmemory-multi-query - Query expansion for better results
• llmemory-multi-tenant - Owner-based isolation for SaaS
• llmemory-rag - Building complete RAG systems

When to use: Building RAG applications, semantic search, document management, or chatbots with knowledge bases.

Key features:

• Hybrid search (vector + BM25 with RRF)
• Query expansion (heuristic + LLM semantic variants)
• Multiple rerankers (OpenAI, CrossEncoder, Lexical)
• Query routing (answerable detection, prevents hallucinations)
• Contextual retrieval (Anthropic's approach)
• Hierarchical chunking with true parent context
• Multi-tenant with owner-based isolation
• pgvector with HNSW indexes
• Multi-language support (14+ languages)
• SOTA RAG compliance (95/100)

# Install all llmemory skills
/plugin install llmemory@juanre-ai-tools

# Or specific skills
/plugin install llmemory-hybrid-search@juanre-ai-tools
/plugin install llmemory-rag@juanre-ai-tools

🤖 llmring - Unified LLM Interface

Single interface for OpenAI, Anthropic, Google Gemini, and Ollama with streaming, tools, and structured output.

Package on GitHub | PyPI

Skills included:

• llmring - Complete skill set (all 6 skills)
• llmring-chat - Basic chat completions ⭐
• llmring-streaming - Streaming responses
• llmring-tools - Function calling and tool use
• llmring-structured - JSON schema and typed responses
• llmring-lockfile - Aliases, profiles, and configuration
• llmring-providers - Multi-provider switching and fallbacks

When to use: Building LLM applications, multi-provider setups, AI agents, or any application integrating with LLMs.