|
| 1 | +--- |
| 2 | +layout: default |
| 3 | +title: Ecosystem |
| 4 | +nav_order: 4 |
| 5 | +permalink: /ecosystem |
| 6 | +--- |
| 7 | + |
| 8 | +# Ecosystem |
| 9 | +{: .no_toc } |
| 10 | + |
| 11 | +The RubyLLM ecosystem includes complementary libraries that extend the core functionality to support advanced use cases like Model Context Protocol (MCP) integration and structured data schemas. |
| 12 | +{: .fs-6 .fw-300 } |
| 13 | + |
| 14 | +## Table of contents |
| 15 | +{: .no_toc .text-delta } |
| 16 | + |
| 17 | +1. TOC |
| 18 | +{:toc} |
| 19 | + |
| 20 | +--- |
| 21 | + |
| 22 | +## RubyLLM::MCP |
| 23 | + |
| 24 | +**Model Context Protocol Support for Ruby** |
| 25 | + |
| 26 | +[RubyLLM::MCP](https://github.com/patvice/ruby_llm-mcp) brings the [Model Context Protocol](https://modelcontextprotocol.io/) to Ruby, enabling your applications to connect to MCP servers and use their tools, resources, and prompts as part of LLM conversations. |
| 27 | + |
| 28 | +### What is MCP? |
| 29 | + |
| 30 | +The Model Context Protocol is an open standard that allows AI applications to integrate with external data sources and tools. MCP servers can expose: |
| 31 | + |
| 32 | +- **Tools**: Functions that LLMs can call to perform actions |
| 33 | +- **Resources**: Structured data that can be included in conversations |
| 34 | +- **Prompts**: Predefined prompt templates with parameters |
| 35 | + |
| 36 | +### Key Features |
| 37 | + |
| 38 | +- 🔌 Multiple transport types (HTTP streaming, STDIO, SSE) |
| 39 | +- 🛠️ Automatic tool integration with RubyLLM |
| 40 | +- 📄 Resource management for files and data |
| 41 | +- 🎯 Prompt templates with arguments |
| 42 | +- 🔄 Support for multiple simultaneous MCP connections |
| 43 | + |
| 44 | +### Installation |
| 45 | + |
| 46 | +```bash |
| 47 | +gem install ruby_llm-mcp |
| 48 | +``` |
| 49 | + |
| 50 | +For detailed documentation, examples, and usage guides, visit the [RubyLLM::MCP documentation](https://rubyllm-mcp.com/). |
| 51 | + |
| 52 | +--- |
| 53 | + |
| 54 | +## RubyLLM::Schema |
| 55 | + |
| 56 | +**Ruby DSL for JSON Schema Creation** |
| 57 | + |
| 58 | +[RubyLLM::Schema](https://github.com/patvice/ruby_llm-schema) provides a clean, Rails-inspired DSL for creating JSON schemas. It's designed specifically for defining structured data schemas for LLM function calling and structured outputs. |
| 59 | + |
| 60 | +### Why Use RubyLLM::Schema? |
| 61 | + |
| 62 | +When working with LLMs, you often need to define precise data structures for: |
| 63 | + |
| 64 | +- Structured output formats |
| 65 | +- Function parameter schemas |
| 66 | +- Data validation schemas |
| 67 | +- API response formats |
| 68 | + |
| 69 | +RubyLLM::Schema makes this easy with a familiar Ruby syntax. |
| 70 | + |
| 71 | +### Key Features |
| 72 | + |
| 73 | +- 📝 Rails-inspired DSL for intuitive schema creation |
| 74 | +- 🎯 Full JSON Schema compatibility |
| 75 | +- 🔧 Support for all primitive types, objects, and arrays |
| 76 | +- 🔄 Union types with `any_of` |
| 77 | +- 📦 Schema definitions and references for reusability |
| 78 | + |
| 79 | +### Installation |
| 80 | + |
| 81 | +```bash |
| 82 | +gem install ruby_llm-schema |
| 83 | +``` |
| 84 | + |
| 85 | +For detailed documentation and examples, visit the [RubyLLM::Schema repository](https://github.com/patvice/ruby_llm-schema). |
| 86 | + |
| 87 | +--- |
| 88 | + |
| 89 | +## Community Projects |
| 90 | + |
| 91 | +The RubyLLM ecosystem is growing! If you've built a library or tool that extends RubyLLM, we'd love to hear about it. Consider: |
| 92 | + |
| 93 | +- Opening a PR to add your project to this page |
| 94 | +- Sharing it in our GitHub Discussions |
| 95 | +- Using the `ruby_llm` topic on your GitHub repository |
| 96 | + |
| 97 | +Together, we're building a comprehensive ecosystem for LLM-powered Ruby applications. |
0 commit comments