Orchestra Docs
This documentation is actively maintained and continuously updated as Orchestra evolves. Features, APIs, and best practices are regularly refined. For the most current information, check back frequently or join our Slack community.
Orchestra is a composable AI agent infrastructure built on LangGraph and powered by the MCP & A2A protocols by Ruska Labs.
What is Orchestra?
Orchestra provides a flexible, API-first platform for building and deploying AI agents. Unlike monolithic AI solutions, Orchestra gives you the freedom to compose your own AI workflows using best-in-class protocols and models.
The goal of Orchestra is to enrich the lives of the curious - those who seek to buy back their time and compound their personal growth. Those who build and don't wait for IT to be built for them.
Who Is This For?
- Developers building AI-powered applications who need flexible, protocol-based integrations
- Teams wanting to leverage multiple AI models and tools in a unified platform
- Organizations seeking to maintain control over their AI infrastructure while using open standards
- Innovators who want to experiment with cutting-edge agent protocols (MCP, A2A)
Core Features
🤖 Flexible AI Assistants
Create pre-configured AI agents with specific instructions, tools, and behaviors. Deploy them via API or use them directly in the web interface.
- Custom Instructions: Define personality and expertise for each assistant
- Tool Integration: Connect assistants to MCP servers, A2A agents, and built-in tools
- Persistent Configuration: Save and reuse assistant configurations across projects
💬 Conversation Threads
Manage stateful conversations with full message history and context persistence.
- Message History: Full conversation tracking with tool execution logs
- File Attachments: Upload and reference documents within threads
- Model Flexibility: Switch between models mid-conversation
- Multi-Modal Support: Text, images, and audio inputs
🔌 Protocol Integration
Orchestra provides native support for open agent protocols, making it easy to connect external tools and services.
- Model Context Protocol (MCP): Connect to MCP-compatible tools and data sources
- Agent-2-Agent (A2A): Enable communication between autonomous agents
- Search Integration: Built-in web search via Searx
Explore Tools & Integrations →
📁 Storage & RAG
S3-compatible object storage with document retrieval capabilities for building knowledge-enhanced agents.
- File Management: Upload and manage documents via S3-compatible API (MinIO)
- RAG Projects: Create searchable document indexes for retrieval-augmented generation
- Flexible Querying: Filter and search across your document collections
Quick Start
Ready to get started? Follow our Getting Started Guide to:
- Access your Orchestra instance
- Create your first conversation thread
- Select and configure AI models
- Integrate external tools via MCP or A2A
Community & Support
- Slack: Join our community for discussions and support
- GitHub: Contribute to Ruska Labs projects
- API Documentation: Complete API reference and interactive docs
- Social Media: Follow us for updates and news
Orchestra is developed by Ruska Labs - enriching the lives of builders and innovators.