ageerle/ruoyi-ai

Features

• Multi-Agent Orchestration Platforms - Coordinates multiple AI agents, integrates model providers, and executes multi-step workflows.
• Agent Workflow Orchestrations - Defines and executes sequences of AI agent tasks that collaborate to complete processes, routing subtasks and merging results.
• Agentic Task Orchestrators - Decomposes complex requests into subtasks managed by a supervisor agent that delegates and coordinates specialized sub-agents.
• AI Model Integrations - Integrates with multiple AI model providers and third-party platforms through a unified interface.
• AI Tooling Frameworks - Ships a typed framework for defining executable tools that AI agents call to interact with external systems.
• AI Workflow Builders - Provides a drag-and-drop interface for designing multi-step AI workflows with tool calls and conditional logic.
• Retrieval-Augmented Generation - Supplements generative outputs with relevant external knowledge retrieved from vector stores or documents.
• Model Provider Abstractions - Abstracts multiple AI provider SDKs behind a common interface, enabling seamless provider switching without changing application code.
• Multi-Agent Orchestrators - Orchestrates specialized AI agents with a supervisor pattern to solve complex tasks together.
• Multi-Model AI Interfaces - Provides a unified abstraction layer for switching between multiple AI model providers without code changes.
• Provider-Agnostic Model Interfaces - Abstracts multiple AI providers behind a unified interface, allowing seamless model swapping and third-party platform connectivity.
• Retrieval Augmented Generation Pipelines - Indexes documents into vector stores and retrieves relevant context at query time to ground generative responses in proprietary data.
• Supervisor Agent Configurations - Coordinates specialized sub-agents through a central supervisor that delegates tasks and merges results for complex collaborative workflows.
• Visual AI Workflow Builders - Ships a drag-and-drop canvas for constructing multi-step AI pipelines executed with real-time streaming output.
• Knowledge Base Retrieval - Parses documents, stores embeddings in a vector database, and retrieves relevant context for AI responses.
• Retrieval Augmented Generation Engines - Indexes internal documents and retrieves relevant context to ground AI responses in proprietary data.
• Multi-Type Step Chains - Defines and executes multi-step workflows that chain tool calls, agents, and conditional logic.
• Tool-Calling Frameworks - Defines a typed interface for agent-invoked external functions, handling parameter validation and result parsing within a unified call framework.
• Tool-Calling Interfaces - Defines executable tools that agents can invoke to interact with external systems via a structured function-calling protocol.
• Custom Tool Exposures - Creates custom tools for AI agents to invoke when interacting with external systems.
• Custom Tool Definitions - Provides a typed protocol for defining executable tools that AI agents invoke to interact with external systems.
• Visual Workflow Engines - Provides a drag-and-drop canvas for constructing multi-step AI pipelines, converting visual diagrams into executable step sequences.
• AI Workflow Designers - Provides a drag-and-drop canvas for constructing multi-step AI pipelines, then executes them with real-time streaming output.

Star history