OxyGent: Build Modular Multi-Agent AI Systems
Summary
OxyGent is a Python framework for assembling tools, language models, and agents into modular multi-agent systems. It suits developers who need reusable components, configurable agent collaboration, and observable execution paths.
At a glance
- Language
- Python
- License
- NOASSERTION
- Stars
- 2.1k
- Forks
- 276
- Added to OSRepos
- March 28, 2026
- Last analyzed
- October 4, 2026
Topics
Click on any tag to explore related repositories
Use at your own risk
OSRepos shares public repositories for knowledge and discovery only. Any installation, execution, configuration, or use of code from these repositories is the user's own responsibility. Always review the repository, source code, dependencies, licenses, and security implications before running or installing anything. OSRepos is not responsible for issues, damages, or losses resulting from third-party repositories.
Overview
OxyGent is a Python framework for building multi-agent applications from standardized components called Oxy, including tools, models, and agents. It aims to make systems easier to assemble and evolve by providing reusable components and interfaces rather than requiring each application to define its own integration patterns.
The framework is a fit for teams that need agents to collaborate on tasks, use tools, or change planning patterns over time. Its README describes dynamic planning, dependency mapping, and auditability as part of its approach. It is less relevant for projects that only need a single model call or a fixed, simple script.
Key Features
- Combines tools, language models, and agents through a shared Oxy component model.
- Supports reusable components and hot-swapping across scenarios.
- Provides ReAct agents and describes dynamic and hybrid planning approaches.
- Lets developers define agent relationships and sub-agents in Python.
- Includes a web service entry point for running a multi-agent system.
- Describes dependency mapping, visual debugging, and decision auditability.
- Includes evaluation engines intended to support feedback and agent improvement.
Use Cases
- Build an assistant that delegates questions to specialist agents for tasks such as time lookup, file operations, and calculations.
- Prototype multi-step task handling where agents need to decompose work and use tools.
- Reuse common tools and agent components across multiple Python applications.
- Explore distributed multi-agent applications when a single-agent workflow is insufficient.
Project Facts
- Language: Python
- License: NOASSERTION
- Stars: 2.1k
- Forks: 276
- Topics: none listed
- Archived: false
Getting Started
Install the package:
pip install oxygent
OxyGent requires Python 3.10 or later and an API key for a compatible LLM provider for the README's example. See the README and official documentation for configuration and usage details.
Alternatives
- Agentarium: Agentarium focuses on context-based agent actions and memory for repeatable multi-agent experiments, while OxyGent emphasizes modular components and observable execution.
- harness-sdk: Strands offers a customizable agent SDK and ready-made harness across Python and TypeScript, rather than focusing on modular multi-agent systems.
- ROMA: ROMA centers on recursive planning to solve complex tasks with coordinated agents, while OxyGent focuses on assembling reusable components into multi-agent systems.
- lagent: Lagent provides synchronous and asynchronous interfaces for composing agents, tools, memory, and workflows; OxyGent emphasizes configurable collaboration and observability.
Considerations
- Applications that use an LLM need provider configuration and credentials. The README also describes compatible API endpoints for local models.
- The README recommends Node.js when using MCP.
- The repository input reports the license as
NOASSERTION, although the README advertises Apache 2.0. Confirm the applicable license in the repository before relying on it. - Claims about distributed scaling and continuous improvement are presented by the project; assess these against your application's requirements before adopting the framework.
Source repository
Open the original repository on GitHub.
8 counted GitHub visits
Related repositories
Similar repositories that may be relevant next.
web-design: Create Consistent Web Pages with a Claude Code Skill
October 3, 2026
web-design is a Claude Code skill that turns product briefs, reference URLs, or screenshots into an editable design specification before generating web code. It is suited to developers and designers who want a repeatable, spec-led workflow for building consistent pages.
oomwoo: Build a DIY Robot Vacuum
October 2, 2026
OOMWOO is a planned, hackable robot vacuum built around Raspberry Pi, ROS2 and 2D LiDAR. It is aimed at makers who want to build and customize a locally controlled vacuum, but its hardware and build instructions are still in development.
shepherd: Supervise Agents with Reversible Execution Traces
October 2, 2026
Shepherd records agent work as inspectable, reversible execution traces and keeps changes as proposals for review. It is aimed at developers building systems that supervise, replay, or manage the work of other agents.
agent-anvil: Test AI Agent Tool Use in CI
October 1, 2026
Agent Anvil evaluates tool-using AI agents through scenario-based runs, trace checks, and optional semantic grading. It helps teams catch unsafe or incorrect tool behavior before it reaches production.