Jakubantalik
Featured
thinking-orbs
A React component library that renders six hand-tuned animated thought orb loading indicators on a plain 2D canvas, with two purpose-tuned sizes and automatic theme detection for AI and agent UIs.
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It demonstrates a structured, real-world setup for customizing a coding agent with consistent rules, MCP integrations, and workflow automation.
Who it is for
- pi coding agent users
- Developers who want to standardize AI-assisted coding workflows
- Teams adopting agent configuration as part of their development environment
- Developers interested in MCP-based tool integrations for coding assistants
Use cases
- Replicating a complete pi agent configuration across multiple machines
- Enforcing Simplified Chinese output and coding conventions during AI-assisted development
- Adding MCP-powered web search, code search, and Chrome DevTools capabilities to a coding agent
- Setting up terminal UI, packages, and custom extensions for pi
Strengths
- Clear directory structure with documented configuration tables
- Includes concrete MCP servers for code search, web search, and Chrome DevTools
- Defines explicit coding principles, workflows, and command strategies in AGENTS.md
- Supports installable packages and custom extensions for extended functionality
Considerations
- Required packages must be installed manually in the terminal
- Configuration is specific to the pi ecosystem and may need adaptation for other agents
- The README does not explain the custom extension source code in detail
README quick start
🧠 my-pi
个人 pi 编码智能体 Agent 配置仓库。
📦 概述
本仓库集中管理 pi Agent 的个性化配置、扩展、MCP 服务以及开发规范，旨在为 AI 编码助手提供一致的行为准则、工具链和工作流。
🗂️ 目录结构
.
├── AGENTS.md # 全局开发配置（语言、编码原则、工作流、命令策略）
├── settings.json # pi 核心设置（主题、包、提供商、代理等）
├── mcp.json # MCP 服务器配置（chrome-devtools、searchcode、tavily）
├── open-tui.json # Open TUI 终端 UI 配置
├── extensions/ # 自定义扩展
│ ├── custom-provider.ts
│ ├── permission-gate.ts
│ ├── qna.ts
│ ├── question.ts
│ ├── questionnaire.ts
│ ├── structured-output.ts
│ └── tools.ts
├── LICENSE
└── README.md
⚙️ 配置说明
settings.json
| 配置项 | 说明 |
|---|---|
theme | 界面主题（one-dark） |
defaultProvider / defaultModel | 默认 AI 提供商与模型 |
httpProxy | HTTP 代理地址 |
packages | 需要额外安装的 pi 包（见下方安装说明） |
retry | 自动重试策略（最多 8 次，基础延迟 5s） |
hideThinkingBlock | 隐藏思考过程 |
quietStartup | 静默启动 |
packages 安装
settings.json 中配置的 packages 需要单独安装，在终端执行：
pi install pi-mcp-adapter
pi install pi-open-tui
pi install @tintinweb/pi-subagents
pi install @pi-lab/notify
pi install @juicesharp/rpiv-todo
pi install @firstpick/pi-themes-bundle
pi install pi-rounded-tools
各包功能：
| 包名 | 说明 |
|---|---|
pi-mcp-adapter | MCP 协议适配器 |
pi-open-tui | 终端 UI 增强 |
@tintinweb/pi-subagents | 子代理支持 |
@pi-lab/notify | 通知功能 |
@juicesharp/rpiv-todo | 任务管理 |
@firstpick/pi-themes-bundle | 主题包合集 |
pi-rounded-tools | 圆角工具样式 |
MCP 服务
mcp.json 配置了以下 MCP 服务器：
- chrome-devtools — Chrome DevTools 协议集成（通过
npx chrome-devtools-mcp） - searchcode — 公共代码搜索与分析
- tavily-remote-mcp — 网络搜索（实时信息、新闻、事实）
Open TUI
open-tui.json 配置了终端 UI 的显示语言、图标、底部状态栏段落和遥测选项。
📜 开发规范（AGENTS.md）
AGENTS.md 是 AI 编码助手的行为准则，核心要点：
- 语言：所有输出使用简体中文
- 编码原则：先思考再编码、简洁优先、外科手术式修改、目标驱动执行
- 工作流：普通功能 → 编码 → 审查 → 提交；复杂功能先规划
- 命令策略：文件操作用专用工具，Git 只读操作自动执行，管理员/交互命令交用户执行
- 自动代理：代码审查
/code-review-expert、复杂规划/planning-with-files自动触发
🚀 快速开始
- 克隆本仓库
- 安装依赖包（见上方
packages 安装章节） - 将
settings.json、mcp.json、open-tui.json放置在 pi 配置目录中 - 将
AGENTS.md放置在项目根目录作为 AI 行为准则
感谢
感谢 LinuxDo 社区对本项目的支持
Description
个人的pi Agent 配置仓库
Related repositories
Similar projects matched by category, topics, and programming language.
makecindy
Featured
cindy
Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.
mereyabdenbekuly-ctrl
Featured
clodex-ide
Clodex is an open-source, local-first agentic IDE that combines persistent AI tasks, code, terminal, browser, Git, models, memory, and governed execution in one Electron workspace, currently in technical preview.