vastsa / PI-Desktop
vastsa / PI-Desktop
PI-Desktop: Your local-first desktop workspace for AI coding agents. Bring your own model. Open any local project. Let agents work — while you stay in control. No PI-Desktop account. No mandatory relay. No editor lock-in. PI-Desktop：专为 AI 编程智能体打造的本地优先桌面工作区。自带模型，打开任意本地项目。让智能体工作，同时保持你的掌控权。无需 PI-Desktop 账户，无强制中继，不绑定特定编辑器。
Download PI-Desktop · Documentation · Screenshots · 简体中文 下载 PI-Desktop · 文档 · 截图 · 简体中文
A standalone desktop workspace for coding agents, projects, models, tools, and long-running sessions. 这是一个为编程智能体、项目、模型、工具和长时间运行的会话提供的独立桌面工作区。
Important: PI-Desktop is currently in Early Preview. The project is actively developed and already usable for real coding workflows, but APIs, extension interfaces, and some desktop behaviors may continue to evolve. 重要提示：PI-Desktop 目前处于早期预览阶段。该项目正在积极开发中，已可用于实际的编程工作流，但 API、扩展接口和部分桌面行为可能会持续演进。
Why PI-Desktop? Most coding agents live inside a terminal, an editor extension, or a hosted service. PI-Desktop gives them a workspace of their own. 为什么选择 PI-Desktop？大多数编程智能体运行在终端、编辑器插件或托管服务中。PI-Desktop 为它们提供了一个专属的工作空间。
🖥️ Desktop-first: Work across repositories and sessions without tying your agent workflow to one editor or terminal. Projects, conversations, reviews, files, previews, notifications, and extensions live in one workspace. 🖥️ 桌面优先：跨仓库和会话工作，无需将智能体工作流绑定到特定的编辑器或终端。项目、对话、审查、文件、预览、通知和扩展程序都集中在一个工作区中。
✨ Bring your own model: Use OpenAI, Anthropic, local models, hosted gateways, or any OpenAI-compatible API. Configure multiple providers and models, then switch between them per session. ✨ 自带模型：使用 OpenAI、Anthropic、本地模型、托管网关或任何兼容 OpenAI 的 API。配置多个提供商和模型，并在每个会话中灵活切换。
🔍 Inspectable by default: Agents can read files, edit code, and run commands — but privileged actions pass through PI-Desktop’s permission layer. Review diffs, inspect command output, and decide how much autonomy each session gets. 🔍 默认可审查：智能体可以读取文件、编辑代码和运行命令，但特权操作必须通过 PI-Desktop 的权限层。你可以审查差异（diff）、检查命令输出，并决定每个会话的自主权限。
🧩 Built to extend: Add Skills, MCP servers, Subagents, and installable Plugins. Plugins can contribute tools, commands, panels, themes, services, skills, and new workspace experiences. 🧩 可扩展设计：添加技能（Skills）、MCP 服务器、子智能体（Subagents）和可安装的插件。插件可以贡献工具、命令、面板、主题、服务、技能以及全新的工作区体验。
From prompt to patch (从提示词到补丁)
Getting started only takes a few steps: 开始使用只需几个步骤：
- Connect a model: Open Settings → Model configuration, choose a provider or compatible API, and add your credentials. 连接模型： 打开设置 → 模型配置，选择提供商或兼容的 API，并添加你的凭据。
- Open a project: Add any local repository or project directory from the sidebar. 打开项目： 从侧边栏添加任何本地仓库或项目目录。
- Pick Agent, Plan, or Goal:
选择 Agent、Plan 或 Goal 模式：
- Agent: starts working. Agent： 直接开始工作。
- Plan: waits until you approve a frozen implementation plan. Plan： 等待你批准一份固定的实施计划。
- Goal: waits until you approve the outcome, then the agent chooses the path. Goal： 等待你批准最终结果，智能体自行选择实现路径。
- Review the result: Inspect edits in the Review panel, check command output, preview the application, and continue the conversation without leaving PI-Desktop. 审查结果： 在审查面板中检查编辑内容、查看命令输出、预览应用程序，无需离开 PI-Desktop 即可继续对话。
More than a chat window (不仅仅是聊天窗口)
Agent, Plan, and Goal: Same agent. Three gates. Privileged tools still go through the permission layer in every mode. Agent、Plan 和 Goal：同一个智能体，三种门控。在所有模式下，特权工具仍需通过权限层。
| Agent | Plan | Goal | |
|---|---|---|---|
| You approve | Nothing extra | The implementation plan | The outcome and acceptance criteria |
| 你批准 | 无需额外操作 | 实施计划 | 结果和验收标准 |
| The agent does | Reads, edits, runs commands, tests, iterates | Studies the repo, writes a frozen plan, then waits | Picks the path and works until the goal is met |
| 智能体执行 | 读取、编辑、运行命令、测试、迭代 | 研究仓库，编写固定计划，然后等待 | 选择路径并工作直到达成目标 |
| Use when | You want it to just do the work | The change is large or risky and you want the approach first | You care about the result, not the route |
| 适用场景 | 你只想让它直接干活 | 变更较大或有风险，你想先确认方案 | 你只关心结果，不关心过程 |
Agent is the default loop: inspect the tree, patch files, run commands, and keep going. Plan is the approval boundary. The agent researches first and produces an immutable implementation plan. Execution does not start until you sign off. Goal is outcome-first. You lock the objective and acceptance criteria; the agent decides how to get there. Agent 是默认循环：检查目录树、修补文件、运行命令并持续进行。Plan 是审批边界，智能体先进行研究并生成不可变的实施计划，在你签字前不会开始执行。Goal 是结果优先，你锁定目标和验收标准，由智能体决定如何达成。
Delegate work to Subagents (将工作委派给子智能体)
Large tasks rarely belong in one context window. PI-Desktop can delegate independent work to background Subagents for things like: codebase exploration, multi-file implementation, research and investigation, test analysis, adversarial review. 大型任务很少能在一个上下文窗口中完成。PI-Desktop 可以将独立工作委派给后台的子智能体，例如：代码库探索、多文件实现、研究与调查、测试分析、对抗性审查。
Each Subagent runs in its own context and reports its result back to the parent agent. 每个子智能体都在自己的上下文中运行，并将结果报告回父智能体。
A workspace built for long sessions (为长会话打造的工作区)
PI-Desktop is designed for more than one prompt at a time. You can manage multiple projects and sessions, pin or archive conversations, branch sessions, queue prompts while an agent is running, reference files with @, use slash commands, and search across the application. PI-Desktop 专为同时处理多个提示词而设计。你可以管理多个项目和会话、固定或归档对话、分支会话、在智能体运行时排队提示词、使用 @ 引用文件、使用斜杠命令，并在整个应用程序中进行搜索。
Streaming responses are checkpointed so interrupted work can survive application restarts or runtime failures whenever possible. 流式响应会进行检查点记录，以便在可能的情况下，中断的工作可以在应用程序重启或运行时故障后恢复。
Your models, your choice (你的模型，你的选择)
PI-Desktop does not lock the agent runtime to a hardcoded model list. Use: OpenAI and Anthropic, OpenAI-compatible APIs, hosted model gateways, local gateways such as Ollama and LM Studio, multiple models under the same provider, provider OAuth accounts where supported. PI-Desktop 不会将智能体运行时锁定在硬编码的模型列表中。你可以使用：OpenAI 和 Anthropic、兼容 OpenAI 的 API、托管模型网关、Ollama 和 LM Studio 等本地网关、同一提供商下的多个模型，以及支持的提供商 OAuth 账户。
Model configuration can include context windows, output limits, reasoning controls, temperature, and other model-specific behavior. Switch models directly from the Composer without recreating your session. 模型配置可包含上下文窗口、输出限制、推理控制、温度（Temperature）以及其他模型特定的行为。无需重新创建会话，即可直接在 Composer 中切换模型。
Review the work, not just the answer (审查工作，而不仅仅是答案)
The agent workspace includes dedicated surfaces for the things that matter while coding: Long-running conversations with transcript navigation, Switch providers, models, and reasoning levels per session, Extend the workspace through the plugin marketplace, Add a provider and connect a model. 智能体工作区包含针对编程关键事项的专用界面：带有记录导航的长会话、按会话切换提供商/模型/推理级别、通过插件市场扩展工作区、添加提供商并连接模型。
Extensions without rebuilding the app (无需重新构建应用即可扩展)
PI-Desktop has several extension layers depending on how deeply you want to customize the agent. PI-Desktop 拥有多个扩展层，取决于你希望对智能体进行多深度的定制。
- Skills: Give agents reusable instructions and workflows. Skills can be installed globally or activated for individual projects. 技能 (Skills)： 为智能体提供可重用的指令和工作流。技能可以全局安装，也可以针对单个项目激活。
- MCP: Connect external tools and services through Model Context Protocol servers without baking them into the desktop application. MCP： 通过模型上下文协议（Model Context Protocol）服务器连接外部工具和服务，无需将其硬编码到桌面应用中。
- Subagents: Create specialized agents with their own instructions, tools, and model choices, then delegate work to them from another agent. 子智能体 (Subagents)： 创建具有独立指令、工具和模型选择的专业智能体，并从另一个智能体委派工作给它们。
- pi extensions: Extensions written for the pi CLI run inside PI-Desktop’s agent unchanged. pi 扩展： 为 pi CLI 编写的扩展可以在 PI-Desktop 的智能体中直接运行，无需修改。