OpenRath
A multi-agent and multi-session workflow runtime similar to PyTorch.
- Type
- MCP
- Open source
- Yes
- GitHub Stars
- ★ 1.1k
- Source
- mcp-github
- Repository
- github.com/Rath-Team/OpenRath
Overview
OpenRath is an open-source, PyTorch-like framework for multi-agent and multi-session workflows. It transforms agent runtime states into explicit, composable Python objects such as Session, Sandbox, Memory, etc. This capability enables AI to collaborate and manage complex multi-agent and multi-session environments. With OpenRath, developers can easily build and manage large-scale multi-agent systems. Ideal for developers working on complex dialogue and collaboration scenarios.
Capabilities
- ▪Supports multi-agent and multi-session
- ▪Session carries conversation state and agent collaboration lineage
- ▪Sandbox determines the actual execution environment for tools
- ▪Memory persists agent state across sessions
- ▪Tool provides model-visible operation interfaces
- ▪Agent acts as a reusable session transformation layer
Use cases
Setup
pip install openrath
This information was compiled by AI from public sources and may contain inaccuracies — please refer to the source.
FAQ
What execution environments does OpenRath support?
Local processes, OpenSandbox, and other backends.
How do I get started with OpenRath?
After installation, read the documentation and refer to example code.
Related skills
gemini-cli
Gemini CLI is an open-source AI agent that brings Gemini's powerful capabilities directly into the terminal.
private-gpt
Provides a complete API layer for local models, supporting RAG, skills, tools, and more.
planning-with-files
File-based persistent planning skill, ideal for AI coding agents and long-running tasks.
scientific-agent-skills
Transform any AI agent into a scientific assistant with 147 ready-to-use research skills.
susi_alexa_skill
A skill that enables question-and-answer interactions between Alexa and Susi AI.
claude-context
Provides code search capabilities for Claude Code, turning the entire codebase into context.