Clioloop — The Open-Source AI Agent with Agentic Fusion

Clioloop — The Open-Source AI Agent with Agentic Fusion

Clioloop — 具备“智能体融合”功能的开源 AI 智能体

We just open-sourced Clioloop, an AI agent with a feature we’re calling Agentic Fusion. I want to share how it works and why we built it. 我们刚刚开源了 Clioloop,这是一个具备我们称之为“智能体融合”(Agentic Fusion)功能的 AI 智能体。我想分享一下它的工作原理以及我们构建它的初衷。

The Problem

核心痛点

Most AI assistants give you one model’s answer. If it’s wrong, you catch it or you don’t. If you use a cheap model, quality drops. If you use a frontier model, you pay frontier prices for everything — even a simple file rename. 大多数 AI 助手只会提供单一模型的回答。如果回答有误,你可能发现,也可能没发现。如果你使用廉价模型,质量就会下降;如果你使用前沿模型,则无论处理什么任务(哪怕只是简单的文件重命名),都要支付高昂的前沿模型费用。

What is Agentic Fusion?

什么是“智能体融合”?

When you run /fusion, a panel of models collaborates on your task: 当你运行 /fusion 命令时,一组模型将协同完成你的任务:

  • Planners (up to 5): Read-only models that research and propose routes in parallel. They figure out the best approach but can’t touch your files or run commands.
    • 规划者(最多 5 个): 只读模型,负责并行研究并提出执行路径。它们负责找出最佳方案,但无法触碰你的文件或运行命令。
  • Main model: Your chosen model does the actual work — full tool access, fully visible. You watch every step. Not a black box.
    • 主模型: 你选择的模型负责实际工作——拥有完整的工具访问权限,且过程完全透明。你可以实时观察每一步操作,它不是一个黑盒。
  • Reviewers (up to 5): Read-only models that critique the draft. They can see images the main model generated. They check for errors, suggest fixes, flag issues.
    • 审查者(最多 5 个): 只读模型,负责对草稿进行评估。它们可以看到主模型生成的图像,负责检查错误、建议修复方案并标记问题。
  • Verdict loop: The draft is revised until reviewers approve. The answer you get has already passed independent review.
    • 结论循环: 草稿会不断修改,直到审查者批准为止。你最终得到的答案已经通过了独立审查。
  • Fusion: Everything combines into one reviewed, approved answer. The quality comes from synthesis — not from running the same job 5 times. Cheap open models combine into something that rivals a frontier model at a fraction of the cost.
    • 融合: 所有内容汇聚成一个经过审查和批准的最终答案。质量源于综合,而不是简单地运行五次相同的任务。廉价的开源模型通过组合,能以极低的成本实现媲美前沿模型的效果。

Safety by Construction

原生安全性

Planners and reviewers are read-only at the schema level. They can research and critique, but they can never touch your files or execute commands. Only your main model has tool access, and you watch it work live. 规划者和审查者在架构层面就是只读的。它们可以进行研究和评估,但永远无法触碰你的文件或执行命令。只有你的主模型拥有工具访问权限,并且你可以实时监控它的工作。

Beyond Fusion

融合之外的功能

Clioloop is also: Clioloop 还具备以下特性:

  • Self-improving: Keeps MEMORY.md and USER.md, updated automatically as it learns your preferences and projects.
    • 自我完善: 维护 MEMORY.mdUSER.md 文件,随着对你偏好和项目的了解自动更新。
  • Autonomous: Set a standing goal with /goal and it loops — planning, running tools, checking progress — until done.
    • 自主性: 通过 /goal 设置长期目标,它会自动循环执行——规划、运行工具、检查进度——直到任务完成。
  • Everywhere: Terminal, desktop app, web dashboard, Telegram, Slack, Discord, WhatsApp — same session, same memory.
    • 全平台覆盖: 终端、桌面应用、Web 面板、Telegram、Slack、Discord、WhatsApp——保持相同的会话和记忆。
  • Multi-agent Kanban: Break big work into a board of tasks; worker agents pick them up and report back.
    • 多智能体看板: 将大型工作拆解为任务看板;工作智能体会领取任务并汇报结果。
  • Tools: File editing, shell, web search & extract, cloud browser, image generation (ComfyUI/Flux), video generation, TTS, any MCP server.
    • 工具集: 文件编辑、Shell、网页搜索与提取、云端浏览器、图像生成(ComfyUI/Flux)、视频生成、TTS 以及任何 MCP 服务器。
  • Scheduled jobs: Run it on cron for automated workflows.
    • 定时任务: 支持 Cron 定时运行,实现自动化工作流。
  • Open-source: Self-host everything. Own your data. No vendor lock-in.
    • 开源: 完全支持自托管。掌控你的数据,无厂商锁定。

The Omni Loop Portal

Omni Loop 门户

One OAuth login gives you access to 300+ models. No API keys to manage. An OpenAI-compatible proxy means you can point any existing tool at it. The tool gateway handles web search, browser automation, image gen, and TTS — all metered, all behind one account. 通过一次 OAuth 登录即可访问 300 多个模型,无需管理 API 密钥。兼容 OpenAI 的代理接口意味着你可以将任何现有工具指向它。工具网关处理网页搜索、浏览器自动化、图像生成和 TTS——所有功能均按量计费,统一账户管理。

Install

安装

macOS / Linux

curl -fsSL https://raw.githubusercontent.com/Clioloop/Clioloop-agent/main/scripts/install.sh | bash

Windows (PowerShell)

iex (irm https://raw.githubusercontent.com/Clioloop/Clioloop-agent/main/install.ps1)

Then: clio setup → pick Omni Loop Portal → approve in browser → start looping 随后执行:clio setup → 选择 Omni Loop Portal → 在浏览器中授权 → 开始循环。

相关链接

We’re building this in the open. Feedback, criticism, feature requests — all welcome. What would you want from an AI agent? 我们正在公开构建这个项目。欢迎任何反馈、批评和功能建议。你希望 AI 智能体具备什么功能?