Skip to content

pnetwork/operator-agent

Repository files navigation

Pentium Agent - [v1.2.0]

Feature in v1.2.0

  • 新增 Mac(Intel) 本版安裝
  • 建立 Github Release 下載連結

Pentium Agent

奔騰網路科技股份有限公司 正式發佈新產品 Pentium Agent 的 BETA 版本。這是一款專為維運及開發人員設計的 AI Agent,BETA 版本可免費下載並安裝於您的 Windows 或 Mac 裝置上。立即開始建立您的個人 AI 助手!

Pentium Agent 提供主流且友善的語言模型聊天介面,輕鬆添加 MCP Server,並可自在的設定 Prompt 以及切換設定,助您輕鬆建立屬於自己的 AI 工具於日常工作流程。主要特色包括:

  1. 主流友善的語言模型介面:簡單直覺的介面,讓您快速與大型語言模型 (LLM) 對話並展開任務。
  2. 輕鬆設定 MCP:無需繁瑣的 Config 檔案,透過直觀的 UI 新增 MCP 伺服器,或從 Pentium 提供的 MCP 市集快速添加 MCP Server。
  3. 自定義 Prompt:Pentium Agent 提供極大的彈性,可以自行定義並且切換。

下載 Pentium Agent

您可以在 Release 內下載最新版本的 Pentium Agent,請確保您的環境已安裝 Docker 環境

目前支援的安裝版本:

  • Windows
  • Mac(Intel)
  • Mac(ARM)

快速開始

啟動 Pentium 應用程式後,您可以依照以下步驟開始使用 Pentium Agent:

  1. 設定 AI 提供者:新增您偏好的語言模型平台。您可以在 LLM 提供商 中找到您的語言模型平台,並在產生 API Key 後,將其新增至 Pentium Agent,並選擇欲使用的模型。
  2. 設定 MCP 伺服器:在 MCP 頁面新增您的 MCP 伺服器,或從 Pentium 市集選擇並加入。Pentium 市集提供的 MCP 伺服器皆以 Docker 方式提供,使用前請確認您的 Docker 環境已啟動。
  3. 與 LLM 聊天並建立任務:在 Pentium Agent 中您可以像使用主流的對話應用程式一樣,與 LLM 進行對話,解決您的日常工作與生活大小事,並建立任務。
  4. 建立日常任務:您可以透過「請幫我建立一個任務」來建立日常任務。提供你要執行的工作排程時間,並描述工作內容,Pentium Agent 將會為您建立任務。
  5. 客製化模型助理:現在你可以自定義模型使用的 Prompt 樣板,自由定義並根據使用情境切換,讓 Pentium Agent 更符合您的使用情境。

社群回饋

Pentium Agent 仍在積極開發中,我們非常重視您的意見與回饋。如果您遇到任何問題或有任何建議,歡迎透過以下管道與我們聯繫:

終端使用者授權合約

EULA


Pentium Agent - [v1.2.0]

Feature in v1.2.0

  • Added Mac(Intel) installation support
  • Created GitHub Release download link

Pentium Agent

Pentium Network Technology Co., Ltd. officially releases the BETA version of its new product Pentium Agent.
This is an AI Agent designed specifically for operations and developers. The BETA version is free to download and can be installed on your Windows or Mac devices. Start building your personal AI assistant today!

Pentium Agent provides a mainstream and user-friendly language model chat interface, making it easy to add MCP Servers, customize prompts, and switch settings freely—helping you effortlessly integrate your own AI tools into daily workflows. Key features include:

  1. Mainstream, user-friendly LLM interface: A simple and intuitive UI that allows you to quickly converse with large language models (LLMs) and start tasks.
  2. Easy MCP configuration: No need for complex config files—add MCP servers via an intuitive UI, or quickly add from the MCP marketplace provided by Pentium.
  3. Customizable Prompts: Pentium Agent offers great flexibility, allowing you to define and switch prompts with ease.

Download Pentium Agent

You can download the latest version of Pentium Agent from the Release page.
Please make sure you have Docker installed in your environment.

Currently supported installation versions:

  • Windows
  • Mac(Intel)
  • Mac(ARM)

Quick Start

After launching Pentium Agent, follow these steps to get started:

  1. Set Up Your AI Provider: Add your preferred LLM provider. You can find supported platforms in the LLM Provider Guide. Once you generate an API key, add it to Pentium Agent and select the model you’d like to use.
  2. Configure Your MCP Server: Go to the MCP tab to add your own server, or select one from the Pentium Marketplace. All servers are provided via Docker—please ensure Docker is running.
  3. Chat with LLM and Create Tasks: Use the chat interface to communicate with the LLM as you would with any messaging app. Solve daily problems or create tasks directly from the conversation.
  4. Create Daily Tasks: Simply ask, “Help me create a task,” specify the schedule and task description, and Pentium Agent will handle the rest.
  5. Customize Your Assistant: Define your own prompt templates and switch them based on different scenarios. Tailor your assistant’s behavior to your unique needs.

Community Feedback

Pentium Agent is still under active development, and we highly value your feedback. If you encounter any issues or have any suggestions, please feel free to contact us through the following channels:

End-User License Agreement

EULA

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Contributors 3

  •  
  •  
  •