Ruby's capable AI runtime
llm.rb is a zero-dependency Ruby runtime for building agentic LLM apps: a single API across 14+ providers, managed tool loops, tools, skills, MCP/A2A clients, streaming callbacks, concurrency strategies and ActiveRecord/Sequel persistence, plus an interactive agent console.
Use Cases
Build agentic Ruby apps with a managed tool loopConnect agents to MCP serversAgent-to-agent (A2A) communicationDefine callable tools as plain Ruby classesTurn markdown files into skills that spawn subagentsStream content, reasoning and tool events with callbacksRun tools concurrently via async, thread, fiber, fork or ractorPersist agent conversations to files or ActiveRecord/SequelInteractive IRB-style console for debugging agentsRAG over local documentsSwitch between 14+ LLM providers behind one APIUse provider-native tools like web search and code execution
Built With
- Language
- Ruby
- Frameworks
- Ruby · ActiveRecord · Sequel · Rack · Rails · async · curb/libcurl · curses · Oj · Yajl · RSpec
Tags
ruby · llm-runtime · agents · tool-calling · mcp-client · a2a · rag · skills · streaming · multi-provider · concurrency · persistence · zero-dependencies · active-record · sequel · console