Claude Skill
stainlu/openclaw-managed-agents
OpenClaw Managed Agents is an open-source, model-agnostic alternative to Claude Managed Agents and ChatGPT Workspace Agents. Supports any model, any cloud, fully transparent.
Overview
Repository
Install this Skill
git clone https://github.com/stainlu/openclaw-managed-agentsRegistry
Summary
OpenClaw Managed Agents is an open-source, model-agnostic alternative to Claude Managed Agents and ChatGPT Workspace Agents. It supports any model and any cloud platform, providing full flexibility and transparency for AI agent orchestration.
OpenClaw 托管代理 —— Claude 托管代理与 ChatGPT 工作区代理的开源替代方案。支持任意模型、任意云平台，完全开源。
Key features
- Open-source alternative to Claude Managed Agents & ChatGPT Workspace Agents
- Model-agnostic: works with any AI model
- Cloud-agnostic: deploy on any cloud platform
- Full transparency and customizability
- Built with TypeScript for reliability and scalability
Use cases
- Building custom AI agent workflows without vendor lock-in
- Deploying multi-model agent systems across different clouds
- Experimenting with open-source agent orchestration for research
- Replacing proprietary managed agents with a transparent solution
README excerpt
# OpenClaw Managed Agents [](https://github.com/stainlu/openclaw-managed-agents/actions/workflows/test.yaml) [](https://github.com/stainlu/openclaw-managed-agents/actions/workflows/publish-images.yaml) [](https://www.npmjs.com/package/@stainlu/openclaw-managed-agents) [](https://pypi.org/project/openclaw-managed-agents/) [](./LICENSE) The open alternative to Claude Managed Agents. Run autonomous AI agents via API — any model, any cloud, open source. Built on [OpenClaw](https://github.com/openclaw/openclaw), the most popular open-source AI agent framework. ## Demo <video src="https://github.com/user-attachments/assets/84a2e1b1-ef6a-4f20-8cbe-4dc36a6c8b77" controls></video> ## What this is **The name says it:** `openclaw-managed-agents` = **OpenClaw** (the agent runtime — multi-provider, 53 built-in skills, MCP-native, Pi's durable JSONL event log) + **Managed Agents** (the four-primitive API shape — Agent / Environment / Session / Event — that Claude Managed Agents made standard). You POST an Agent (model + system prompt + tools + MCP servers), open a Session against it, send Events, stream back Events. Under the hood: one isolated Docker container per active session running OpenClaw, SQLite for orchestrator metadata, SSE for streaming, WebSocket control plane for cancel / model-override / tool-confirmation. Session state is durable even if the live container is la
Topics
No topics yet.