
Agentspan是开源AI智能体工作流运行时,支持崩溃恢复、人工审批、安全防护和全链路可观测性。MIT协议,轻松部署您现有的Agent框架。

Agentspan 是一个开源服务器和 SDK,为 AI 智能体提供持久化执行能力。你的智能体定义会编译成弹性工作流——执行状态保存在进程外部,工具调用自动重试,人工审批可以暂停并干净地恢复,即使你的进程崩溃也无妨。
>> 将智能体连接成持久化管道。每一步都有日志、防崩溃、可在整个链路上恢复。mock_run 精确编排工具序列,实现确定性 CI 测试——无需 LLM 调用,无需服务器,毫秒级执行。Agentspan 面向将 AI 智能体投入生产的工程团队——需要超越演示框架可靠性的开发者、寻求 Netflix 级编排的基础设施工程师,以及需要在智能体系统中实现审计追踪、人工监督和跨会话记忆的组织。

Open source, free, local debugger for AI agents.

Ship AI agents without the operational burden

A local control plane for AI coding agents

Build coding agents with a plugin-based open-source runtime

open source agent engineering platform

An open source AI harness built with the human in mind

Shared AI memory that stops agents from repeating mistakes

Control AI agents with confidence

The Infrastructure Behind AI Agencies | White-Label Platform

The Self-Improvement Stack For agents

Open-Source Brain For Your Team

build your own software factory

Cron jobs for AI agents w/ one command, zero infrastructure

OpenRouter for agent tools

Your Chief Agent Operator for multi-agent work

AI Meeting companion with cross-meeting memory

An open source AI harness built with the human in mind

An AI wearable that remembers your conversations all day

Al sleep companion that helps fall asleep without struggle

Skip the prompting. Produce consistently compelling videos.
AI agents that turn signals into crypto + Polymarket trades
The scraping service AI agents run on

Predict the next Series A from a ProductHunt launch

Your AI director for creating cinematic videos with ease

See where Claude Code burns tokens. Hit your limits less.