# Agentao

> MIT governed Python agent runtime — permissions, MCP, memory, replay

**Category:** AI Agents  
**Pricing:** open-source  
**URL:** https://agentao.cn  
**GitHub:** https://github.com/jin-bo/agentao  
**Added:** 2026-09-23

---

## The Hook

Agentao is a local-first governed agent runtime you embed in Python or run as CLI/ACP — permissions, protocols, memory, plugins, and audit replay are first-class.

## What's Good

MIT (jin-bo/agentao; `pip install agentao`). Embeddable harness for hosts that need permissioned tools and MCP, not a free-for-all REPL. Developer guide at agentao.cn (EN/ZH); explicit embed-for-agents playbooks for Claude Code/Codex. Beta on PyPI.

## The Catch

Vs Pi, Unreal Agent, and LangGraph: Agentao is a governed Python embed/CLI runtime, not a TS coding harness, durable-ops library, or graph orchestration framework. Docs and community skew bilingual/China-first; verify the embed API against your host before betting a production path on beta.

## Verdict

MIT governed Python agent runtime when permissions, MCP, and replay matter inside your app — not a drop-in Claude Code clone.