qianshou 3.0.3

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Qianshou Contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,278 @@
+<h1 align="center">千手</h1>
+
+<p align="center">
+  <strong>面向办公人士的 AI 智能体应用。</strong><br>
+  理解项目文档、代码库。规划复杂任务。自主执行。<br>
+  你的终端 AI 队友。
+</p>
+
+---
+
+## 目录
+
+- [什么是 千手](#什么是-千手)
+- [核心功能](#核心功能)
+- [支持的模型](#支持的模型)
+- [安装](#安装)
+- [快速开始](#快速开始)
+- [使用方法](#使用方法)
+- [实验性功能](#实验性功能)
+
+---
+
+## 什么是千手
+
+千手 是一款运行在终端中的**智能体编码工具**。与传统的代码补全工具不同，千手 能够理解你的整个代码库，规划复杂的多步骤任务，并自主执行。
+
+把它当作一个 AI 队友，它可以：
+- 阅读并理解大规模项目
+- 将复杂需求分解为可执行的步骤
+- 编写、修改和重构跨多个文件的代码
+- 运行终端命令并管理 Git 工作流
+- 解释代码逻辑并生成文档
+
+---
+
+## 核心功能
+
+### 🧠 智能体编码
+
+千手 作为一个**自主智能体**运行，而不仅仅是一个被动的助手。当你给它一个高层次的目标，比如"添加 OAuth 用户认证"时，它会：
+
+1. **理解意图** - 分析你的自然语言请求
+2. **扫描上下文** - 阅读相关文件并理解项目结构
+3. **制定计划** - 将任务分解为清晰、可执行的步骤
+4. **执行** - 在你的批准下实施每个步骤
+5. **迭代** - 报告进度并在需要时寻求澄清
+
+### 📚 代码库级理解
+
+支持 **200K+ token 上下文窗口**，千手 可以一次性分析整个项目。它理解：
+- 项目结构和模块依赖关系
+- 编码模式和设计约定
+- 跨文件关系和架构决策
+
+### 🔧 深度代码分析
+
+- **解释复杂代码** - 获取函数、算法和逻辑流的详细说明
+- **追踪函数调用** - 跟踪代码库中的执行路径
+- **识别问题** - 检测性能瓶颈、安全漏洞和潜在错误
+- **智能重构** - 执行大规模、跨文件的代码转换
+
+### ⚡ 工作流自动化
+
+千手 与你的开发工作流无缝集成：
+
+- **运行终端命令** - 执行 `npm install`、运行测试、构建 Docker 镜像
+- **Git 集成** - 搜索提交历史、解决合并冲突、生成提交信息
+- **测试生成** - 自动创建全面的单元测试和集成测试
+- **文档生成** - 生成 API 文档、代码注释和 README 文件
+
+### 🔌 可扩展架构
+
+- **MCP 协议** - 通过模型上下文协议连接外部工具和 API
+- **自定义技能** - 编写自己的命令和工作流
+- **插件系统** - 通过社区插件扩展功能
+- **IDE 无关** - 适用于任何编辑器（VS Code、JetBrains、Vim 等）
+
+---
+
+## 支持的模型
+
+千手 原生支持多个 AI 提供商。设置相应的环境变量即可切换提供商。
+
+常用配置：
+```bash
+export QS_AUTH_TOKEN="..."
+export QS_BIG_MODEL=""
+export QS_MIDDLE_MODEL=""
+export QS_SMALL_MODEL=""
+export API_TIMEOUT_MS=600000
+export QS_BASE_URL="https://coding.dashscope.aliyuncs.com/apps/anthropic"
+export QS_BIG_MODEL= ...
+#直接运行即可， auto-run 是无需再对每个命令进行确认。如果对于非常严肃的场景，需要去掉 auto-run，逐个动作进行审核。
+cd xxx 进入你的目录
+qianshou --auto-run
+
+# 如果要运行桌面办公版，处理一些文档 文件之类
+QS_OFFICE=1
+qianshou --auto-run
+...
+```
+
+你可创建一个 AGENTS.md 文件便于 AI 了解整个项目。
+
+# 智谱
+```bash
+export QS_AUTH_TOKEN="..."
+export QS_BIG_MODEL="glm-4.7"
+export QS_MIDDLE_MODEL="glm-4.7"
+export QS_SMALL_MODEL="glm-4.7"
+export QS_BASE_URL="https://open.bigmodel.cn/api/anthropic"
+qianshou --auto-run
+...
+```
+
+# DeepSeek
+```bash
+export QS_BASE_URL=https://api.deepseek.com/anthropic
+export QS_AUTH_TOKEN=sk-...
+export API_TIMEOUT_MS=600000
+export QS_MODEL=deepseek-chat
+export QS_SMALL_FAST_MODEL=deepseek-chat
+export QS_DISABLE_NONESSENTIAL_TRAFFIC=1
+
+qianshou --auto-run
+```
+
+### OpenAI Codex
+
+使用 OpenAI 的 Codex 模型进行代码生成。
+
+```bash
+export QS_USE_OPENAI=1
+qianshou
+```
+
+### AWS Bedrock
+
+通过你的 AWS 账户路由请求。
+
+```bash
+export QS_USE_AWS=1
+export AWS_REGION="us-east-1"
+qianshou
+```
+
+### Google Cloud Vertex AI
+
+通过你的 GCP 项目路由请求。
+
+```bash
+export QS_USE_VERTEX=1
+qianshou
+```
+
+### Anthropic Foundry
+
+使用 Anthropic Foundry 进行专用部署。
+
+```bash
+export QS_USE_CLAUDE_FOUNDRY=1
+export ANTHROPIC_FOUNDRY_API_KEY="..."
+qianshou
+```
+
+---
+
+## 安装
+
+### 快速安装
+
+npm i -g @zytech/qianshou
+ 
+
+### 系统要求
+
+- **运行时**: [Bun](https://bun.sh) >= 1.3.11
+- **操作系统**: macOS、Linux 或 Windows（通过 WSL）
+- **认证**: 所选提供商的 API 密钥或 OAuth 登录
+
+---
+
+## 快速开始
+
+```bash
+# 使用你喜欢的提供商进行身份验证
+./qianshou /login
+
+# 启动交互式 REPL
+./qianshou
+
+# 单次执行模式
+./qianshou -p "这个目录里有什么文件？"
+
+# 指定模型
+./qianshou --model ...
+```
+
+---
+
+## 使用方法
+
+### 交互式 REPL
+
+```bash
+./qianshou --auto-run
+```
+
+### 对话示例
+
+**实现新功能：**
+```
+你: 添加一个用户注册的 REST API 端点
+
+Qianshou: 我来帮你添加用户注册端点。让我先检查你的项目结构...
+
+[阅读文件，分析代码库]
+
+我将创建以下内容：
+1. POST /api/users/register 端点
+2. 用户验证模式
+3. 用户的数据库模型
+4. 端点的单元测试
+
+我可以继续吗？
+```
+
+**调试问题：**
+```
+你: 上传文件时出现 500 错误。这是堆栈跟踪...
+
+Qianshou: 查看错误和你的代码，问题似乎出现在文件上传处理程序中。
+问题在第 42 行，缓冲区大小处理不当。让我修复这个问题...
+```
+
+**重构代码：**
+```
+你: 重构认证模块以使用 JWT 令牌
+
+Qianshou: 我将重构你的认证模块以使用 JWT。这需要：
+1. 安装 jsonwebtoken 包
+2. 创建 JWT 工具函数
+3. 更新登录/登出处理程序
+4. 添加令牌验证中间件
+
+[继续重构]
+```
+
+---
+
+## 实验性功能
+
+`build:dev:full` 构建启用了 54 个实验性功能标志。亮点包括：
+
+### 交互与 UI
+- **ULTRAPLAN** - 远程多智能体规划
+- **ULTRATHINK** - 用于复杂推理的深度思考模式
+- **VOICE_MODE** - 按住说话语音输入
+- **TOKEN_BUDGET** - 使用跟踪和警告
+- **HISTORY_PICKER** - 交互式提示历史
+- **QUICK_SEARCH** - 快速提示搜索
+
+### 智能体、记忆与规划
+- **BUILTIN_EXPLORE_PLAN_AGENTS** - 内置智能体预设
+- **VERIFICATION_AGENT** - 任务验证
+- **AGENT_TRIGGERS** - 后台自动化
+- **EXTRACT_MEMORIES** - 自动记忆提取
+- **TEAMMEM** - 团队记忆文件
+
+### 工具与基础设施
+- **BRIDGE_MODE** - IDE 远程控制桥接
+- **BASH_CLASSIFIER** - 智能权限决策
+- **MCP_RICH_OUTPUT** - 增强的 MCP 输出
+
+
+---
+
+**用 ❤️ 构建，为想要从 AI 工具中获得更多功能的开发者服务。**

package/README_en.md

@@ -0,0 +1,277 @@
+<h1 align="center">Qianshou</h1>
+
+<p align="center">
+  <strong>AI Agent Application for Office Professionals.</strong><br>
+  Understand project documentation and codebases. Plan complex tasks. Execute autonomously.<br>
+  Your terminal AI teammate.
+</p>
+
+---
+
+## Table of Contents
+
+- [What is Qianshou](#what-is-qianshou)
+- [Core Features](#core-features)
+- [Supported Models](#supported-models)
+- [Installation](#installation)
+- [Quick Start](#quick-start)
+- [Usage](#usage)
+- [Experimental Features](#experimental-features)
+
+---
+
+## What is Qianshou
+
+Qianshou is an **agent-based coding tool** that runs in your terminal. Unlike traditional code completion tools, Qianshou can understand your entire codebase, plan complex multi-step tasks, and execute them autonomously.
+
+Think of it as an AI teammate that can:
+- Read and understand large-scale projects
+- Break down complex requirements into executable steps
+- Write, modify, and refactor code across multiple files
+- Run terminal commands and manage Git workflows
+- Explain code logic and generate documentation
+
+---
+
+## Core Features
+
+### 🧠 Agent-Based Coding
+
+Qianshou runs as an **autonomous agent**, not just a passive assistant. When you give it a high-level goal like "add OAuth user authentication," it will:
+
+1. **Understand Intent** - Analyze your natural language request
+2. **Scan Context** - Read relevant files and understand project structure
+3. **Create Plan** - Break down the task into clear, actionable steps
+4. **Execute** - Implement each step with your approval
+5. **Iterate** - Report progress and seek clarification when needed
+
+### 📚 Codebase-Level Understanding
+
+With support for **200K+ token context windows**, Qianshou can analyze entire projects at once. It understands:
+- Project structure and module dependencies
+- Coding patterns and design conventions
+- Cross-file relationships and architectural decisions
+
+### 🔧 Deep Code Analysis
+
+- **Explain Complex Code** - Get detailed explanations of functions, algorithms, and logic flows
+- **Trace Function Calls** - Follow execution paths through the codebase
+- **Identify Issues** - Detect performance bottlenecks, security vulnerabilities, and potential bugs
+- **Smart Refactoring** - Execute large-scale, cross-file code transformations
+
+### ⚡ Workflow Automation
+
+Qianshou seamlessly integrates with your development workflow:
+
+- **Run Terminal Commands** - Execute `npm install`, run tests, build Docker images
+- **Git Integration** - Search commit history, resolve merge conflicts, generate commit messages
+- **Test Generation** - Automatically create comprehensive unit and integration tests
+- **Documentation Generation** - Generate API docs, code comments, and README files
+
+### 🔌 Extensible Architecture
+
+- **MCP Protocol** - Connect external tools and APIs via Model Context Protocol
+- **Custom Skills** - Write your own commands and workflows
+- **Plugin System** - Extend functionality with community plugins
+- **IDE Agnostic** - Works with any editor (VS Code, JetBrains, Vim, etc.)
+
+---
+
+## Supported Models
+
+Qianshou natively supports multiple AI providers. Simply set the appropriate environment variables to switch providers.
+
+Common configuration:
+```bash
+export QS_AUTH_TOKEN="..."
+export QS_BIG_MODEL=""
+export QS_MIDDLE_MODEL=""
+export QS_SMALL_MODEL=""
+export API_TIMEOUT_MS=600000
+export QS_BASE_URL="https://coding.dashscope.aliyuncs.com/apps/anthropic"
+export QS_BIG_MODEL= ...
+# Run directly, auto-run means no need to confirm each command. For very serious scenarios, remove auto-run and review each action.
+cd xxx # Enter your directory
+qianshou --auto-run
+
+# If you want to run the desktop office version for processing documents and files
+QS_OFFICE=1
+qianshou --auto-run
+...
+```
+
+You can create an AGENTS.md file to help AI understand the entire project.
+
+# Zhipu AI
+```bash
+export QS_AUTH_TOKEN="..."
+export QS_BIG_MODEL="glm-4.7"
+export QS_MIDDLE_MODEL="glm-4.7"
+export QS_SMALL_MODEL="glm-4.7"
+export QS_BASE_URL="https://open.bigmodel.cn/api/anthropic"
+qianshou --auto-run
+...
+```
+
+# DeepSeek
+```bash
+export QS_BASE_URL=https://api.deepseek.com/anthropic
+export QS_AUTH_TOKEN=sk-...
+export API_TIMEOUT_MS=600000
+export QS_MODEL=deepseek-chat
+export QS_SMALL_FAST_MODEL=deepseek-chat
+export QS_DISABLE_NONESSENTIAL_TRAFFIC=1
+
+qianshou --auto-run
+```
+
+### OpenAI Codex
+
+Use OpenAI's Codex models for code generation.
+
+```bash
+export QS_USE_OPENAI=1
+qianshou
+```
+
+### AWS Bedrock
+
+Route requests through your AWS account.
+
+```bash
+export QS_USE_AWS=1
+export AWS_REGION="us-east-1"
+qianshou
+```
+
+### Google Cloud Vertex AI
+
+Route requests through your GCP project.
+
+```bash
+export QS_USE_VERTEX=1
+qianshou
+```
+
+### Anthropic Foundry
+
+Use Anthropic Foundry for dedicated deployments.
+
+```bash
+export QS_USE_CLAUDE_FOUNDRY=1
+export ANTHROPIC_FOUNDRY_API_KEY="..."
+qianshou
+```
+
+---
+
+## Installation
+
+### Quick Install
+
+npm i -g @zytech/qianshou
+
+### System Requirements
+
+- **Runtime**: [Bun](https://bun.sh) >= 1.3.11
+- **Operating System**: macOS, Linux, or Windows (via WSL)
+- **Authentication**: API keys or OAuth login for your chosen provider
+
+---
+
+## Quick Start
+
+```bash
+# Authenticate with your preferred provider
+./qianshou /login
+
+# Start the interactive REPL
+./qianshou
+
+# One-shot execution mode
+./qianshou -p "What files are in this directory?"
+
+# Specify model
+./qianshou --model ...
+```
+
+---
+
+## Usage
+
+### Interactive REPL
+
+```bash
+./qianshou --auto-run
+```
+
+### Conversation Examples
+
+**Implementing a new feature:**
+```
+You: Add a REST API endpoint for user registration
+
+Qianshou: I'll help you add a user registration endpoint. Let me first check your project structure...
+
+[Reading files, analyzing codebase]
+
+I'll create the following:
+1. POST /api/users/register endpoint
+2. User validation schema
+3. Database model for users
+4. Unit tests for the endpoint
+
+Shall I proceed?
+```
+
+**Debugging an issue:**
+```
+You: I'm getting a 500 error when uploading files. Here's the stack trace...
+
+Qianshou: Looking at the error and your code, the issue appears to be in the file upload handler.
+The problem is on line 42, where the buffer size is not handled properly. Let me fix this...
+```
+
+**Refactoring code:**
+```
+You: Refactor the authentication module to use JWT tokens
+
+Qianshou: I'll refactor your authentication module to use JWT. This will involve:
+1. Installing the jsonwebtoken package
+2. Creating JWT utility functions
+3. Updating login/logout handlers
+4. Adding token verification middleware
+
+[Proceeding with refactor]
+```
+
+---
+
+## Experimental Features
+
+The `build:dev:full` build enables 54 experimental feature flags. Highlights include:
+
+### Interaction & UI
+- **ULTRAPLAN** - Remote multi-agent planning
+- **ULTRATHINK** - Deep thinking mode for complex reasoning
+- **VOICE_MODE** - Push-to-talk voice input
+- **TOKEN_BUDGET** - Usage tracking and warnings
+- **HISTORY_PICKER** - Interactive prompt history
+- **QUICK_SEARCH** - Quick prompt search
+
+### Agents, Memory & Planning
+- **BUILTIN_EXPLORE_PLAN_AGENTS** - Built-in agent presets
+- **VERIFICATION_AGENT** - Task verification
+- **AGENT_TRIGGERS** - Background automation
+- **EXTRACT_MEMORIES** - Automatic memory extraction
+- **TEAMMEM** - Team memory files
+
+### Tools & Infrastructure
+- **BRIDGE_MODE** - IDE remote control bridge
+- **BASH_CLASSIFIER** - Intelligent permission decisions
+- **MCP_RICH_OUTPUT** - Enhanced MCP output
+
+
+---
+
+**Built with ❤️ for developers who want more from their AI tools.**

package/package.json

@@ -0,0 +1,126 @@
+{
+  "name": "qianshou",
+  "version": "3.0.3",
+  "description": "Qianshou - AI-powered assistant",
+  "type": "module",
+  "packageManager": "bun@1.3.11",
+  "bin": {
+    "qianshou": "./qianshou.js"
+  },
+  "files": [
+    "qianshou.js",
+    "README.md",
+    "README_en.md",
+    "LICENSE"
+  ],
+  "engines": {
+    "bun": ">=1.3.11"
+  },
+  "scripts": {
+    "build": "bun run ./scripts/build.ts",
+    "build:dev": "bun run ./scripts/build.ts --dev",
+    "build:dev:full": "bun run ./scripts/build.ts --dev --feature-set=dev-full",
+    "compile": "bun run ./scripts/build.ts --compile",
+    "dev": "bun run ./src/entrypoints/cli.tsx"
+  },
+  "dependencies": {
+    "@alcalzone/ansi-tokenize": "^0.3.0",
+    "@anthropic-ai/bedrock-sdk": "^0.26.4",
+    "@anthropic-ai/claude-agent-sdk": "^0.2.87",
+    "@anthropic-ai/foundry-sdk": "^0.2.3",
+    "@anthropic-ai/mcpb": "^2.1.2",
+    "@anthropic-ai/sandbox-runtime": "^0.0.44",
+    "@anthropic-ai/sdk": "^0.80.0",
+    "@anthropic-ai/vertex-sdk": "^0.14.4",
+    "@aws-sdk/client-bedrock": "^3.1020.0",
+    "@aws-sdk/client-bedrock-runtime": "^3.1020.0",
+    "@aws-sdk/client-sts": "^3.1020.0",
+    "@aws-sdk/credential-provider-node": "^3.972.28",
+    "@aws-sdk/credential-providers": "^3.1020.0",
+    "@azure/identity": "^4.13.1",
+    "@commander-js/extra-typings": "^14.0.0",
+    "@growthbook/growthbook": "^1.6.5",
+    "@modelcontextprotocol/sdk": "^1.29.0",
+    "@opentelemetry/api": "^1.9.1",
+    "@opentelemetry/api-logs": "^0.214.0",
+    "@opentelemetry/core": "^2.6.1",
+    "@opentelemetry/exporter-logs-otlp-grpc": "^0.214.0",
+    "@opentelemetry/exporter-logs-otlp-http": "^0.214.0",
+    "@opentelemetry/exporter-logs-otlp-proto": "^0.214.0",
+    "@opentelemetry/exporter-metrics-otlp-grpc": "^0.214.0",
+    "@opentelemetry/exporter-metrics-otlp-http": "^0.214.0",
+    "@opentelemetry/exporter-metrics-otlp-proto": "^0.214.0",
+    "@opentelemetry/exporter-prometheus": "^0.214.0",
+    "@opentelemetry/exporter-trace-otlp-grpc": "^0.214.0",
+    "@opentelemetry/exporter-trace-otlp-http": "^0.214.0",
+    "@opentelemetry/exporter-trace-otlp-proto": "^0.214.0",
+    "@opentelemetry/resources": "^2.6.1",
+    "@opentelemetry/sdk-logs": "^0.214.0",
+    "@opentelemetry/sdk-metrics": "^2.6.1",
+    "@opentelemetry/sdk-trace-base": "^2.6.1",
+    "@opentelemetry/semantic-conventions": "^1.40.0",
+    "@smithy/core": "^3.23.13",
+    "@smithy/node-http-handler": "^4.5.1",
+    "ajv": "^8.18.0",
+    "asciichart": "^1.5.25",
+    "auto-bind": "^5.0.1",
+    "axios": "^1.14.0",
+    "bidi-js": "^1.0.3",
+    "cacache": "^20.0.4",
+    "chalk": "^5.6.2",
+    "chokidar": "^5.0.0",
+    "cli-boxes": "^4.0.1",
+    "cli-highlight": "^2.1.11",
+    "code-excerpt": "^4.0.0",
+    "diff": "^8.0.4",
+    "emoji-regex": "^10.6.0",
+    "env-paths": "^4.0.0",
+    "execa": "^9.6.1",
+    "fflate": "^0.8.2",
+    "figures": "^6.1.0",
+    "fuse.js": "^7.1.0",
+    "get-east-asian-width": "^1.5.0",
+    "google-auth-library": "^10.6.2",
+    "highlight.js": "^11.11.1",
+    "https-proxy-agent": "^8.0.0",
+    "ignore": "^7.0.5",
+    "indent-string": "^5.0.0",
+    "ink": "^6.8.0",
+    "jsonc-parser": "^3.3.1",
+    "lodash-es": "^4.17.24",
+    "lru-cache": "^11.2.7",
+    "marked": "^17.0.5",
+    "p-map": "^7.0.4",
+    "picomatch": "^4.0.4",
+    "plist": "^3.1.0",
+    "proper-lockfile": "^4.1.2",
+    "qrcode": "^1.5.4",
+    "react": "^19.2.4",
+    "react-reconciler": "^0.33.0",
+    "semver": "^7.7.4",
+    "sharp": "^0.34.5",
+    "shell-quote": "^1.8.3",
+    "signal-exit": "^4.1.0",
+    "stack-utils": "^2.0.6",
+    "strip-ansi": "^7.2.0",
+    "supports-hyperlinks": "^4.4.0",
+    "tree-kill": "^1.2.2",
+    "turndown": "^7.2.2",
+    "type-fest": "^5.5.0",
+    "undici": "^7.24.6",
+    "usehooks-ts": "^3.1.1",
+    "vscode-jsonrpc": "^8.2.1",
+    "vscode-languageserver-protocol": "^3.17.5",
+    "vscode-languageserver-types": "^3.17.5",
+    "wrap-ansi": "^10.0.0",
+    "ws": "^8.20.0",
+    "xss": "^1.0.15",
+    "xxhash-wasm": "^1.1.0",
+    "yaml": "^2.8.3",
+    "zod": "^4.3.6"
+  },
+  "devDependencies": {
+    "@types/bun": "^1.3.11",
+    "typescript": "^6.0.2"
+  }
+}