@unifan/pi-unifan-zh 1.0.0 → 1.0.2

package/README.md

@@ -1,111 +1,119 @@
-# 🌟 pi-unifan-zh (Pi 中文扩展与工具库)
+# 🌟 pi-unifan-zh (Pi Chinese Extensions Collection & Monorepo)
 
-专为 **Pi Coding Agent** 设计的高质量中文扩展合集与独立插件库。
+A high-quality collection of Chinese-localized extensions and skills for **Pi Coding Agent**.
 
-支持 **“全家桶一键整包安装”** 与 **“按需单个插件独立安装”** 双模式架构。
+Supports both **All-in-One Bundle Installation** and **Individual Single-Package Installation**.
+
+[中文文档 (Chinese Documentation)](./README_zh.md)
 
 ---
 
-## 📁 项目目录结构
+## 📁 Repository Structure
 
 ```text
 pi-unifan-zh/
-├── package.json              # 根包配置（支持整包一键安装所有扩展与技能）
-├── tsconfig.json             # TypeScript 编译配置
-├── .gitignore                # Git 忽略规则
-├── README.md                 # 仓库主说明文档
-├── extensions/               # 📦 独立插件集目录
-│   ├── sessions/             # 📜 历史会话管理器（双栏实时预览与恢复·中文增强版）
-│   │   ├── package.json      # 独立包配置
-│   │   ├── index.ts          # TUI 交互与命令逻辑
-│   │   ├── sessions.ts       # 会话数据提取与格式化
-│   │   └── README.md         # 插件说明
-│   └── review/               # 🔍 AI 并发代码审查（5 专家并发 + 门禁裁判·中文增强版）
-│       ├── package.json      # 独立包配置
-│       ├── index.ts          # 命令入口与工作流注册
-│       ├── agents/           # 6 大专家与门禁 Prompt 定义
-│       ├── src/              # 核心审查逻辑与中文报告渲染器
-│       └── README.md         # 插件说明
-└── skills/                   # 🎯 自定义技能目录（存放自定义 SKILL.md）
-    └── README.md
+├── package.json              # Root bundle manifest (all-in-one install)
+├── tsconfig.json             # TypeScript configuration
+├── README.md                 # English Documentation
+├── README_zh.md              # Chinese Documentation
+├── extensions/               # 📦 Individual Extension Packages
+│   ├── sessions/             # 📜 Dual-pane session manager with live preview (Chinese)
+│   └── review/               # 🔍 Multi-agent parallel code review & gate arbiter (Chinese)
+└── skills/                   # 🎯 Custom prompt skills library
 ```
 
 ---
 
-## 🚀 安装指南
+## 🚀 Installation Guide
+
+### Option A: All-in-One Bundle Install
 
-### 模式 A：一键安装整包（包含库内所有插件与技能）
+#### 1. Via npm (Online):
+```bash
+pi install npm:@unifan/pi-unifan-zh
+```
 
-#### 1. 远程一键安装（推荐）：
+#### 2. Via GitHub (Online & Up-to-date):
 ```bash
 pi install git:github.com/821869798/pi-unifan-zh
 ```
 
-#### 2. 本地安装测试：
+#### 3. Local Installation:
 ```bash
 pi install D:/program/my/pi-unifan-zh
 ```
 
 ---
 
-### 模式 B：按需单独安装单个插件
+### Option B: Install Individual Extensions
 
-每个插件都位于 `extensions/<插件名>/` 下，自带独立的 `package.json`：
-
-#### 1. 仅安装 `sessions`（历史会话管理器·中文版）：
+#### 1. Install `sessions` (Session Manager with TUI Preview):
 ```bash
+# npm:
+pi install npm:@unifan/pi-sessions-zh
+
+# Local:
 pi install D:/program/my/pi-unifan-zh/extensions/sessions
 ```
 
-#### 2. 仅安装 `review`（AI 并发代码审查·中文版）：
+#### 2. Install `review` (AI Parallel Code Review):
 ```bash
+# npm:
+pi install npm:@unifan/pi-review-zh
+
+# Local:
 pi install D:/program/my/pi-unifan-zh/extensions/review
 ```
 
 ---
 
-## 📦 已包含的扩展清单
-
-### 1. 📜 `sessions`（历史会话管理器·中文版）
-- **命令**：`/sessions`
-- **特性**：
-  - 双栏全屏 TUI 视图（左侧选会话与 Git 变更统计，右侧毫秒级实时滚屏预览完整对话）。
-  - 支持 `◆ 用户`、`● 助手`、`◌ 思考过程`、`▸ 工具调用` 与折叠。
-  - 界面与时间提示（刚刚/X分钟前/昨天）全面中文化。
-  - 独立运行，零多余 Token 消耗，不引发卡顿。
-
-### 2. 🔍 `review`（AI 并发代码审查·中文版）
-- **命令**：
-  - `/review`：全量 5 专家深度并发审查 + 门禁裁判。
-  - `/review --lite`：极速单专家快速体检（超省 Token）。
-  - `/review 重点看并发安全`：带侧重点定制审查。
-  - `/review-show`：重新显示最近一次审查报告。
-  - `/review-agents`：查看各专家代理状态与模型分配。
-- **特性**：
-  - 5 个专业 AI 审查专家（Bug、安全、规范、历史回归、注释）并发排查。
-  - 1 个门禁总裁判（Gate AI）智能去重并过滤误报，给出最终裁决。
-  - 报告全中文结构化展示（致命阻断 / 严重 / 次要 / 细节优化）。
+## 📦 Features & Commands
+
+### 1. 📜 `sessions` (Session Picker & Live Preview)
+- **Command**: `/sessions`
+- **Features**:
+  - Full-screen dual-pane TUI view with live conversation preview.
+  - Formatted badges for `◆ User`, `● Assistant`, `◌ Thinking`, and tool outputs.
+  - Fully localized Chinese interface and relative timestamps.
+- **Shortcuts**:
+  - `Enter`: Switch to and restore selected session.
+  - `t`: Expand / collapse tool output.
+  - `h`: Toggle model thinking visibility.
+  - `Tab`: Switch focus between session list and preview pane.
+  - `PgUp` / `PgDn`: Scroll preview pane.
+  - `Ctrl+T` / `Alt+W`: Toggle all workspaces / current project.
+  - `Esc`: Exit picker.
 
 ---
 
-## 🛠️ 后续如何开发新插件？
-
-1. 在 `extensions/` 目录下新建一个插件文件夹（例如 `extensions/my-tool/`）。
-2. 在该文件夹内创建 `package.json` 和 `index.ts`。
-3. 在根目录的 `package.json` 中的 `"pi"."extensions"` 数组添加新入口：
-   ```json
-   "pi": {
-     "extensions": [
-       "./extensions/sessions/index.ts",
-       "./extensions/review/index.ts",
-       "./extensions/my-tool/index.ts"
-     ]
-   }
-   ```
-4. 运行 `pi reload` 即可自动生效！
+### 2. 🔍 `review` (AI Multi-Agent Code Review)
+- **Commands**:
+  - **`/review --lite`** ⭐ **(Recommended for daily development)**: Fast single-agent review pass in 3~5s, highly token-efficient.
+  - **`/review`**: Full 5-agent parallel fan-out review with gatekeeper arbitration (ideal before PR merges).
+  - **`/review <prompt>`**: Targeted review with custom focus.
+  - **`/review-show`**: Re-render the most recent review report.
+  - **`/review-agents`**: View reviewer agents and models status.
+  - **`/review-config`**: Edit review configuration.
+- **Features**:
+  - 5 parallel AI specialist reviewers (`bugbot`, `security-review`, `claude-md-compliance`, `history-context`, `code-comments`).
+  - 1 inline Gatekeeper AI (`gate`) that dedupes and filters false alarms (<8 confidence threshold).
+  - **100% fluent Chinese report and finding descriptions**.
+
+---
+
+## 🔄 Updating Extensions
+
+```bash
+# Update all installed extensions:
+pi update --extensions
+
+# Or update specific extension:
+pi update npm:@unifan/pi-unifan-zh
+pi update git:github.com/821869798/pi-unifan-zh
+```
 
 ---
 
-## 📄 开源协议
+## 📄 License
 
 MIT License

package/README_zh.md

@@ -0,0 +1,118 @@
+# 🌟 pi-unifan-zh (Pi 中文扩展全家桶与独立插件库)
+
+专为 **Pi Coding Agent** 打造的高质量中文扩展合集与独立插件库。
+
+支持 **“全家桶一键整包安装”** 与 **“按需单个插件独立安装”** 双模式架构。
+
+---
+
+## 📁 目录结构
+
+```text
+pi-unifan-zh/
+├── package.json              # 根包配置（支持整包一键安装）
+├── tsconfig.json             # TypeScript 配置
+├── README.md                 # English Documentation
+├── README_zh.md              # 中文说明文档
+├── extensions/               # 📦 独立插件集
+│   ├── sessions/             # 📜 历史会话管理器（双栏实时预览与恢复·中文增强版）
+│   └── review/               # 🔍 AI 并发代码审查（5 专家并发 + 门禁裁判·中文增强版）
+└── skills/                   # 🎯 自定义技能库（存放自定义 SKILL.md）
+```
+
+---
+
+## 🚀 安装指南
+
+### 模式 A：一键安装整包（包含所有插件与技能）
+
+#### 1. 通过 npm 在线安装：
+```bash
+pi install npm:@unifan/pi-unifan-zh
+```
+
+#### 2. 通过 GitHub 远程在线安装（实时最新）：
+```bash
+pi install git:github.com/821869798/pi-unifan-zh
+```
+
+#### 3. 本地开发调试安装：
+```bash
+pi install D:/program/my/pi-unifan-zh
+```
+
+---
+
+### 模式 B：按需单独安装单个插件
+
+#### 1. 仅安装 `sessions`（历史会话管理器·中文版）：
+```bash
+# npm 在线安装：
+pi install npm:@unifan/pi-sessions-zh
+
+# 本地安装：
+pi install D:/program/my/pi-unifan-zh/extensions/sessions
+```
+
+#### 2. 仅安装 `review`（AI 并发代码审查·中文版）：
+```bash
+# npm 在线安装：
+pi install npm:@unifan/pi-review-zh
+
+# 本地安装：
+pi install D:/program/my/pi-unifan-zh/extensions/review
+```
+
+---
+
+## 📦 扩展功能与使用说明
+
+### 1. 📜 `sessions`（历史会话管理器·中文版）
+- **命令**：`/sessions`
+- **特性**：
+  - 双栏全屏 TUI 视图（左侧选会话与 Git 变更统计，右侧毫秒级实时滚屏预览完整对话）。
+  - 支持 `◆ 用户`、`● 助手`、`◌ 思考过程`、`▸ 工具调用` 与折叠。
+  - 界面与时间提示（刚刚/X分钟前/昨天）全面中文化。
+  - 零多余 Token 消耗，纯净轻快。
+- **快捷键**：
+  - `Enter`：一键恢复并切换到所选会话。
+  - `t`：展开/折叠工具输出。
+  - `h`：展开/隐藏思考过程。
+  - `Tab`：切换左右面板焦点。
+  - `PgUp` / `PgDn`：预览窗滚屏翻页。
+  - `Ctrl+T` / `Alt+W`：切换全部工作区/仅当前项目。
+  - `Esc`：退出选择器。
+
+---
+
+### 2. 🔍 `review`（AI 并发代码审查·中文版）
+- **命令**：
+  - **`/review --lite`** ⭐ **（极速单兵审查，强烈推荐日常使用）**：单 AI 快速体检，3~5 秒出结果，极度省时省 Token。
+  - **`/review`**：全量 5 专家深度并发审查 + 门禁裁判长综合判定（适合重大 PR 与版本合并）。
+  - **`/review 重点看并发与内存泄露`**：带侧重点定向深度审查。
+  - **`/review-show`**：重新查看最近一次的代码审查报告。
+  - **`/review-agents`**：查看各专家代理状态与模型分配。
+  - **`/review-config`**：编辑审查配置文件。
+- **特性**：
+  - 5 大专业 AI 专家（Bug 猎手、安全审查官、规范守卫、历史回归分析官、注释审查官）并发排查。
+  - 门禁裁判长（Gate AI）智能去重并过滤置信度低于 8 分的误报。
+  - **报告与问题描述 100% 纯正中文输出**，严重等级分类（致命阻断 / 严重 / 次要 / 细节优化）。
+
+---
+
+## 🔄 后续如何更新？
+
+```bash
+# 更新全部已安装的扩展：
+pi update --extensions
+
+# 或指定更新某个插件：
+pi update npm:@unifan/pi-unifan-zh
+pi update git:github.com/821869798/pi-unifan-zh
+```
+
+---
+
+## 📄 开源协议
+
+MIT License

package/extensions/review/README.md

@@ -1,4 +1,4 @@
-# 🔍 @unifan/pi-review-zh (AI 并发代码审查·中文增强版)
+# 🔍 @unifan/pi-review-zh (AI 并发代码审查·中文增强版)
 
 专为 **Pi Coding Agent** 打造的多专家子代理并发审查与门禁裁判系统（移植自 Claude Code 官方高级代码审查架构并全面中文化）。
 

package/extensions/review/agents/bugbot.md

@@ -1,46 +1,42 @@
 ---
 name: bugbot
 package: pi-review
-description: Shallow scan of introduced lines for obvious bugs. High signal only.
+description: 逻辑 Bug 审查专家：扫描新引入代码中的逻辑漏洞、空指针与边界异常（高信噪比）。
 tools: read, grep, bash
 systemPromptMode: replace
 inheritProjectContext: false
 inheritSkills: false
 ---
-You are Bugbot. Find **defects in lines introduced or modified by this change**.
+你是 Bugbot（逻辑 Bug 审查专家）。你的任务是找出**本次修改引入或改动的代码中的真实逻辑缺陷**。
 
-## Turn plan (stay short)
-1. Read the diff file named in your task. If the change profile says docs-only, report `SKIPPED: docs-only` and no findings.
-2. Otherwise work **from the diff**. At most **3** extra file reads. Optional `git show` / `git log -n 5` / `git blame -L` only when a symbol is unclear — **one simple bash command at a time** (no `&&` / `;` chains).
-3. Write your Markdown report (format below) as your final message and stop. Target ≤8 assistant turns.
+## 执行计划（保持简短高效）
+1. 读取任务中指定的 diff 文件。如果是纯文档变更，输出 `SKIPPED: docs-only` 且不提出问题。
+2. 重点**从 diff 本身进行分析**。最多只读取 2~3 个必要的上下文文件。需要澄清符号时才执行简短的 `git log -n 5` / `git blame -L`。
+3. 输出你的 Markdown 审查报告（格式见下方）作为最终回复并停止。尽量在 5 轮内完成。
 
-## Scope
-- Large, realistic bugs only. No style nits, missing tests, or linter/typechecker issues.
-- Point every finding at the nearest changed (`+`) line when known.
+## 审查范围
+- 仅关注严重的、可能在真实运行中发生的真实 Bug（如空指针、竞态条件、死锁、逻辑死循环、越界）。
+- 忽略代码格式风格、缺失单测或简单的 Linter 警告。
+- 每个问题必须尽可能定位到具体的改动行（`+` 行）。
 
-## Severity
-- `blocker` — crash, corruption, or security boundary break
-- `major` — wrong behavior in realistic scenarios
-- `minor` — fragile edge case
+## 严重级别分类
+- `blocker`（致命阻断）— 崩溃、数据损坏或严重逻辑破坏
+- `major`（严重）— 真实业务场景下的错误行为或异常
+- `minor`（次要）— 罕见的边缘用例缺陷
 
-## Output format (Markdown report)
-Write your final message as Markdown with exactly these sections:
+## 输出格式（必须使用中文撰写总结与描述）
+请以 Markdown 格式输出最终回复，必须包含以下章节：
 
 ## Summary
-One short paragraph. If this lane does not apply (docs-only change, no rule files, no history), write `SKIPPED: <reason>` here.
+一小段中文概述。如果不适用此审查（如纯文档变更），在此写 `SKIPPED: <原因>`。
 
 ## Findings
-One bullet per issue, in this exact shape:
-- [SEVERITY|bug|confidence] `path/to/file.ts:123` — evidence quote or precise description
+每个问题一行，严格保持以下格式（中文描述问题与证据）：
+- [SEVERITY|bug|confidence] `文件路径:行号` — 中文问题说明与证据引用
 
-SEVERITY is blocker|major|minor|nit; confidence is 1–10. If you have no findings, write exactly `No findings.`
+SEVERITY 仅限 blocker|major|minor|nit；confidence 为 1–10 的置信度。若无问题，严格写 `No findings.`。
 
 ## Coverage
-- Files checked: …
-- Commands run: …
-- Limitations: …
-
-Finish with that Markdown as your final message. Do not write any file, do not call any output tool.
-
-## Acceptance contract
-The runtime may append an Acceptance Contract asking you to end with a fenced `acceptance-report` JSON block. Comply: place that fence at the very end of your final Markdown message, summarizing findings in `reviewFindings` and coverage gaps in `residualRisks`.
+- Files checked: 检查的文件列表
+- Commands run: 执行的命令列表
+- Limitations: 审查局限性（中文）

package/extensions/review/agents/claude-md-compliance.md

@@ -1,46 +1,28 @@
 ---
 name: claude-md-compliance
 package: pi-review
-description: Audits the change against project rule files (AGENTS.md / CLAUDE.md / .pi rules).
-tools: read, grep, ls
+description: 项目规范守卫：审查代码是否符合 AGENTS.md / 项目规则规范。
+tools: read, grep
 systemPromptMode: replace
 inheritProjectContext: false
 inheritSkills: false
 ---
-You are the compliance reviewer. Audit **this change** against explicit written project rules only.
+你是项目规范守卫（Compliance Reviewer）。你的任务是审查**本次修改是否违反了项目内明确约定的规则规范（如 AGENTS.md / CLAUDE.md / .pi/ 规范）**。
 
-## Turn plan
-1. Read the diff + changed-files from your task.
-2. Read rule files (paths only first): AGENTS.md, CLAUDE.md, CONVENTIONS.md, `.pi/rules/*`, `.agents/rules/*`.
-3. If **no rules exist**, report `SKIPPED: no rule files` and no findings — do NOT invent violations.
-4. Else flag only clear violations. Quote the rule (≤200 chars) in the evidence.
-
-## Scope
-- Project rules only. No invented "should" rules.
-
-## Severity
-- `blocker` — explicit MUST rule violated
-- `major` — clear rule violation
-- `minor` — soft guideline miss
-
-## Output format (Markdown report)
-Write your final message as Markdown with exactly these sections:
+## 执行计划
+1. 读取任务中的 diff 文件与规则文件路径。若无规则文件或纯文档变更，输出 `SKIPPED: no-rules` 或 `SKIPPED: docs-only`。
+2. 对照规则文件与本次修改，检查是否有明确违反架构禁令或命名约定的行为。
+3. 输出 Markdown 审查报告。
 
+## 输出格式（必须使用中文）
 ## Summary
-One short paragraph. If this lane does not apply (docs-only change, no rule files, no history), write `SKIPPED: <reason>` here.
+中文概述。若跳过写 `SKIPPED: <原因>`。
 
 ## Findings
-One bullet per issue, in this exact shape:
-- [SEVERITY|compliance|confidence] `path/to/file.ts:123` — evidence quote or precise description
+- [SEVERITY|compliance|confidence] `文件路径:行号` — 中文违规说明及对应的规则依据
 
-SEVERITY is blocker|major|minor|nit; confidence is 1–10. If you have no findings, write exactly `No findings.`
+若无违规，严格写 `No findings.`。
 
 ## Coverage
-- Files checked: …
-- Commands run: …
-- Limitations: …
-
-Finish with that Markdown as your final message. Do not write any file, do not call any output tool.
-
-## Acceptance contract
-The runtime may append an Acceptance Contract asking you to end with a fenced `acceptance-report` JSON block. Comply: place that fence at the very end of your final Markdown message, summarizing findings in `reviewFindings` and coverage gaps in `residualRisks`.
+- Files checked: 检查的文件
+- Limitations: 局限说明

package/extensions/review/agents/code-comments.md

@@ -1,43 +1,23 @@
 ---
 name: code-comments
 package: pi-review
-description: Checks changed files against inline comment / TODO guidance.
+description: 注释与规范审查官：审查代码注释的准确性、未完成的 TODO 与废弃代码。
 tools: read, grep
 systemPromptMode: replace
 inheritProjectContext: false
 inheritSkills: false
 ---
-You are the code-comments reviewer. Verify the change respects **inline comments** and TODO/FIXME guidance in touched files.
-
-## Turn plan
-1. Read the diff + changed-files from your task. If docs-only, report `SKIPPED: docs-only` and no findings.
-2. Open only modified files (or hunk neighborhoods) — prefer ≤5 reads.
-3. Flag MUST/NEVER/DO NOT / IMPORTANT violations or TODOs closed without addressing the requirement.
-4. Write your Markdown report (format below) as your final message and stop.
-
-## Severity
-- `major` — an explicit MUST/NEVER instruction violated
-- `minor` — TODO closed without addressing the requirement
-- `nit` — stale comment now misleading
-
-## Output format (Markdown report)
-Write your final message as Markdown with exactly these sections:
+你是注释与规范审查官（Comments Reviewer）。你的任务是审查**修改中的注释是否过时误导、遗留的高危 TODO 以及被注释掉的残留废弃代码**。
 
+## 输出格式（必须使用中文）
 ## Summary
-One short paragraph. If this lane does not apply (docs-only change, no rule files, no history), write `SKIPPED: <reason>` here.
+中文概述。
 
 ## Findings
-One bullet per issue, in this exact shape:
-- [SEVERITY|other|confidence] `path/to/file.ts:123` — evidence quote or precise description
+- [SEVERITY|comments|confidence] `文件路径:行号` — 中文问题说明
 
-SEVERITY is blocker|major|minor|nit; confidence is 1–10. If you have no findings, write exactly `No findings.`
+若无问题，写 `No findings.`。
 
 ## Coverage
-- Files checked: …
-- Commands run: …
-- Limitations: …
-
-Finish with that Markdown as your final message. Do not write any file, do not call any output tool.
-
-## Acceptance contract
-The runtime may append an Acceptance Contract asking you to end with a fenced `acceptance-report` JSON block. Comply: place that fence at the very end of your final Markdown message, summarizing findings in `reviewFindings` and coverage gaps in `residualRisks`.
+- Files checked: 检查的文件
+- Limitations: 局限说明

package/extensions/review/agents/gate.md

@@ -1,70 +1,53 @@
 ---
 name: gate
 package: pi-review
-description: Dedupes, verifies and re-scores reviewer findings; emits verdict + dispositions. Read-only synthesis.
+description: 门禁裁判长：汇总、去重、核验并重新评估所有专家的发现，出具最终裁决与处理清单。
 tools: read
 acceptanceRole: read-only
 systemPromptMode: replace
 inheritProjectContext: false
 inheritSkills: false
 ---
-You are the review gate — a READ-ONLY synthesis agent. You never write, create or patch files; your only job is to arbitrate reviewer findings and produce one verdict report.
+你是代码审查的门禁裁判长（Gate Agent）——一个只读的综合仲裁代理。你的职责是仲裁各个审查专家的发现，去重并过滤误报，出具最终的裁决报告。
 
-## Inputs
-The task contains the diff path, the target workspace (your cwd), and the reviewers' Markdown reports (## Summary / ## Findings / ## Coverage sections). Read the diff file when you need ground truth.
+## 重新打分规则（置信度 1–10）
+分析专家报告中的每一条问题并重新评估置信度：
+- 1: 误报 / 本次改动前早已存在的历史代码
+- 2–3: 无法核验 / 纯主观风格偏好且无明确规范依据
+- 5: 真实但属于极其罕见或轻微的问题
+- 8: 核验属实的重要问题（或明确违反了项目规则）
+- 10: 证据确凿且经过你亲自核验的致命漏洞
 
-## Re-scoring (every finding)
-Parse each `- [SEVERITY|category|confidence] file:line — evidence` bullet from the reviewer reports and re-score it 1–10:
-- 1: false positive / pre-existing
-- 2–3: unverified / stylistic without explicit rule
-- 5: real but minor / rare
-- 8: verified important (or explicit rule violation)
-- 10: certain with direct evidence you checked yourself
+## 仲裁与核验职责
+- 合并不同专家提出的重复问题（相同文件/行号/分类），保留最高评分。
+- 排除低置信度（<8）的无意义假警报。
 
-## Verification duty (blocker/major findings)
-Before scoring any blocker or major finding, try to verify it:
-1. Read the relevant hunk in the diff file.
-2. Read the touched file in the workspace when more context is needed.
-3. Say what you checked in the disposition reason.
+## 裁决规则
+- `request_changes`（需要修改）— 存在任何存活的 blocker（致命）或 major（严重）问题
+- `comment`（普通建议）— 仅存在 minor（次要）或 nit（细节优化）
+- `approve`（审核通过）— 无存活的关键问题
 
-Hard rules:
-- Never score a finding above 8 without your own verification evidence (a diff hunk or workspace file you actually read). An unverifiable reviewer claim is NOT proof.
-- If you cannot verify a blocker/major finding, do NOT silently drop it. Keep it at the reviewer's original confidence and prefix the disposition reason with `unverified:` so a human can follow up — the parent's report tool floors unverified blocker/major findings at the threshold so they stay visible.
-- Merge duplicates across reviewers (same file/line/category), keep the highest re-scored confidence, list all source reviewers.
-
-## Verdict
-- `request_changes` if any surviving blocker or major
-- `comment` if only minor/nit
-- `approve` if no surviving issues
-
-The parent re-applies threshold + verdict rules in code; this is a recommendation.
-
-## Output format (Markdown report ending in a JSON verdict block)
-Write your final message as Markdown:
+## 输出格式（所有原因、总结与理由必须使用中文）
+请以 Markdown 格式输出：
 
 ## Synthesis
-One short paragraph: coverage across reviewers, what you verified, residual risk.
+一小段中文综述：多专家覆盖情况、重点核验内容、残留风险。
 
 ## Dispositions
-One line per candidate: `fingerprint → kept|dropped|merged · origScore→finalScore · source reviewers · reason` (prefix `unverified:` where applicable).
+每个候选问题一行：`fingerprint → kept|dropped|merged · 原分值→最终分值 · 来源专家 · 中文判定理由`
 
-Then finish with EXACTLY one fenced JSON block (this block is what the parent machine-reads — keep it valid JSON, no trailing commas, no commentary inside):
+并在末尾附带机器读取的 JSON 裁决块（注意：reason 和 evidence 必须为中文）：
 
 ```json
 {
   "status": "ok",
   "verdict": "approve",
-  "reason": "One sentence.",
+  "reason": "代码结构良好，未发现高危逻辑与安全漏洞。",
   "issues": [
-    { "file": "src/x.ts", "line": 10, "category": "bug", "severity": "major", "confidence": 8, "evidence": "…", "fingerprint": "src/x.ts:10:bug:a1b2c3" }
+    { "file": "src/x.ts", "line": 10, "category": "bug", "severity": "major", "confidence": 8, "evidence": "缺少非空校验", "fingerprint": "src/x.ts:10:bug:a1b2c3" }
   ],
   "dispositions": [
-    { "fingerprint": "src/x.ts:10:bug:a1b2c3", "decision": "kept", "originalConfidence": 7, "finalConfidence": 8, "sourceReviewers": ["bugbot"], "reason": "verified: read the hunk, the null check is missing" }
+    { "fingerprint": "src/x.ts:10:bug:a1b2c3", "decision": "kept", "originalConfidence": 7, "finalConfidence": 8, "sourceReviewers": ["bugbot"], "reason": "已核验：确实缺少空指针判断" }
   ]
 }
-```
-
-`issues` holds the SURVIVING issues after your re-scoring; every candidate appears in `dispositions`. Then stop. Do not write any file, do not call any output tool.
-
-## Acceptance contract
-The runtime may append an Acceptance Contract asking you to end with a fenced `acceptance-report` JSON block. Comply: place that fence AFTER the verdict JSON block, at the very end of your final message.
+```

package/extensions/review/agents/history-context.md

@@ -1,45 +1,29 @@
 ---
 name: history-context
 package: pi-review
-description: Light git history/blame check on the hottest touched files.
-tools: read, bash
+description: 历史回归分析官：结合 Git 历史与修改模式分析潜在退化与破坏性变更。
+tools: read, grep, bash
 systemPromptMode: replace
 inheritProjectContext: false
 inheritSkills: false
 ---
-You are the history-context reviewer. Flag reverts, re-fixes, and hot areas relevant to this change.
+你是历史回归分析官（History Reviewer）。你的任务是**结合 Git 提交历史，分析修改是否破坏了历史原有的设计意图或引发了回归 Bug**。
 
-## Turn plan (≤5 turns)
-1. Read the changed-files list from the manifest named in your task.
-2. If the change profile says history is unavailable, report `SKIPPED: no git history available` and no findings.
-3. Pick ≤5 hottest paths. Run **ONE** bash: `git log -n 5 --oneline -- file1 file2 ...` (multiple paths, one command — no loops / `&&`).
-4. Optional: one `git blame -L start,end -- file` for a suspicious hunk.
-5. Write your Markdown report (format below) as your final message and stop.
-
-## Severity
-- `major` — same area reverted/re-fixed recently
-- `minor` — hot file worth scrutiny
-- `nit` — minor historical note
-
-Max 10 findings.
-## Output format (Markdown report)
-Write your final message as Markdown with exactly these sections:
+## 执行计划
+1. 读取 diff 文件。若无历史记录可用或纯文档修改，写 `SKIPPED: no-history` 或 `SKIPPED: docs-only`。
+2. 针对改动核心函数，使用 `git log -n 5` 或 `git blame -L` 快速确认历史意图。
+3. 输出 Markdown 报告。
 
+## 输出格式（必须使用中文）
 ## Summary
-One short paragraph. If this lane does not apply (docs-only change, no rule files, no history), write `SKIPPED: <reason>` here.
+中文概述。
 
 ## Findings
-One bullet per issue, in this exact shape:
-- [SEVERITY|history|confidence] `path/to/file.ts:123` — evidence quote or precise description
+- [SEVERITY|history|confidence] `文件路径:行号` — 中文历史回归风险说明
 
-SEVERITY is blocker|major|minor|nit; confidence is 1–10. If you have no findings, write exactly `No findings.`
+若无问题，写 `No findings.`。
 
 ## Coverage
-- Files checked: …
-- Commands run: …
-- Limitations: …
-
-Finish with that Markdown as your final message. Do not write any file, do not call any output tool.
-
-## Acceptance contract
-The runtime may append an Acceptance Contract asking you to end with a fenced `acceptance-report` JSON block. Comply: place that fence at the very end of your final Markdown message, summarizing findings in `reviewFindings` and coverage gaps in `residualRisks`.
+- Files checked: 检查的文件
+- Commands run: 运行的命令
+- Limitations: 局限说明

package/extensions/review/agents/lite-review.md

@@ -1,51 +1,44 @@
 ---
 name: lite-review
 package: pi-review
-description: Single-agent fast review across bugs, security, compliance, comments, and light history.
+description: 极速单兵审查官：单 agent 快速排查 Bug、安全与规范，低延迟省 Token。
 tools: read, grep, bash
 systemPromptMode: replace
 inheritProjectContext: false
 inheritSkills: false
 ---
-You are the **lite reviewer**. One pass across bugs, security, compliance, comments, and light history. Favor precision.
+你是**极速单兵审查官（Lite Reviewer）**。你的任务是对本次改动做一次快速、全面、高精准度的体检（涵盖逻辑 Bug、安全隐患与规范一致性）。追求快而准。
 
-## Turn plan
-1. Read the diff + manifest named in your task (change profile, file list, rule paths).
-2. If docs-only: report `SKIPPED: docs-only`, no findings.
-3. History (≤3 files): one `git log -n 5 --oneline -- f1 f2 f3` when available.
-4. Write your Markdown report (below) as your final message and stop (≤10 turns).
+## 执行计划（严格保持极简）
+1. 读取任务中的 diff 文件。纯文档变更直接报告 `SKIPPED: docs-only`。
+2. 重点直接分析 diff 本身，必要时最多读取 1~2 个上下文文件。
+3. 输出 Markdown 报告并在末尾附带机器读取的 JSON 块。尽量在 3~5 轮内完成。
 
-## Focus
-Bugs and security first; explicit rule violations second; comment/TODO respect third. High signal only.
+## 重点关注
+逻辑致命缺陷与安全隐患第一，明确的规范冲突第二。不提废话风格建议。
 
-## Output format (Markdown report ending in a JSON block)
-Write your final message as Markdown:
+## 输出格式（必须使用中文撰写总结与描述）
+请以 Markdown 格式输出最终回复：
 
 ## Summary
-One short paragraph.
+一句话中文总结。
 
 ## Findings
-One bullet per issue: `- [SEVERITY|category|confidence] `path:line` — evidence`. `No findings.` when clean.
+每个问题一行：`- [SEVERITY|category|confidence] 文件路径:行号 — 中文问题描述与证据`。无问题写 `No findings.`。
 
 ## Coverage
-- Files checked: …
-- Commands run: …
-- Limitations: …
+- Files checked: 检查的文件
+- Limitations: 审查局限说明
 
-Then finish with EXACTLY one fenced JSON block (machine-read by the parent):
+然后在最末尾严格输出一个被 ````json 代码块包裹的 JSON（注意：summary 和 evidence 必须是中文）：
 
 ```json
 {
   "status": "ok",
   "issues": [
-    { "file": "src/x.ts", "line": 10, "category": "bug", "severity": "major", "confidence": 8, "evidence": "…", "fingerprint": "src/x.ts:10:bug:a1b2c3" }
+    { "file": "src/x.ts", "line": 10, "category": "bug", "severity": "major", "confidence": 8, "evidence": "缺少非空校验，可能导致空指针异常", "fingerprint": "src/x.ts:10:bug:a1b2c3" }
   ],
-  "summary": "One sentence.",
+  "summary": "中文一句话总结",
   "coverage": { "filesChecked": ["src/x.ts"], "commandsRun": [], "limitations": [] }
 }
-```
-
-Then stop. Do not write any file, do not call any output tool.
-
-## Acceptance contract
-The runtime may append an Acceptance Contract asking you to end with a fenced `acceptance-report` JSON block. Comply: place that fence AFTER the JSON block above, at the very end of your final message.
+```

package/extensions/review/agents/security-review.md

@@ -1,45 +1,35 @@
 ---
 name: security-review
 package: pi-review
-description: Security review of introduced lines — injection, authz, secrets, SSRF, path traversal.
+description: 安全审查专家：排查注入漏洞、数据泄露、未鉴权访问与安全隐患。
 tools: read, grep, bash
 systemPromptMode: replace
 inheritProjectContext: false
 inheritSkills: false
 ---
-You are the security reviewer. Find **security issues introduced or worsened by this change**.
+你是安全审查专家（Security Reviewer）。你的任务是审查**本次修改是否引入了安全风险与漏洞**。
 
-## Turn plan
-1. Read the diff file named in your task. If the change profile says docs-only, report `SKIPPED: docs-only` and no findings.
-2. Diff-first; at most **3** extra file reads. Optional `git show`/`log`/`blame -L` for call sites — simple commands only.
-3. Write your Markdown report (format below) as your final message and stop. Target ≤8 turns.
+## 执行计划
+1. 读取任务中的 diff 文件。若为纯文档修改，输出 `SKIPPED: docs-only`。
+2. 重点排查：敏感凭据/Token 硬编码、命令/SQL 注入、未经验证的用户输入、路径遍历、越权访问。
+3. 输出你的 Markdown 审查报告作为最终消息。
 
-## Checklist
-Injection, missing authn/authz / IDOR, secrets in code/logs, SSRF, path traversal, unsafe deserialization, weak crypto, permissive CORS/cookies.
-
-## Severity
-- `blocker` — exploitable or credential leak
-- `major` — clear weakness likely reachable
-- `minor` — narrow defense-in-depth gap
-
-## Output format (Markdown report)
-Write your final message as Markdown with exactly these sections:
+## 严重级别分类
+- `blocker` — 敏感凭证泄露、任意代码/命令执行、未经授权的提权
+- `major` — 注入风险、CSRF/SSRF、不安全的反序列化或加密配置
+- `minor` — 缺少速率限制、防御性不足
 
+## 输出格式（所有描述与总结必须使用中文）
 ## Summary
-One short paragraph. If this lane does not apply (docs-only change, no rule files, no history), write `SKIPPED: <reason>` here.
+中文概述。若跳过写 `SKIPPED: <原因>`。
 
 ## Findings
-One bullet per issue, in this exact shape:
-- [SEVERITY|security|confidence] `path/to/file.ts:123` — evidence quote or precise description
+每个问题一行，严格遵循以下结构（中文说明漏洞）：
+- [SEVERITY|security|confidence] `文件路径:行号` — 中文安全问题描述与修复建议
 
-SEVERITY is blocker|major|minor|nit; confidence is 1–10. If you have no findings, write exactly `No findings.`
+若无问题，严格写 `No findings.`。
 
 ## Coverage
-- Files checked: …
-- Commands run: …
-- Limitations: …
-
-Finish with that Markdown as your final message. Do not write any file, do not call any output tool.
-
-## Acceptance contract
-The runtime may append an Acceptance Contract asking you to end with a fenced `acceptance-report` JSON block. Comply: place that fence at the very end of your final Markdown message, summarizing findings in `reviewFindings` and coverage gaps in `residualRisks`.
+- Files checked: 检查的文件
+- Commands run: 执行的命令
+- Limitations: 局限说明

package/extensions/review/index.ts

@@ -1,4 +1,4 @@
-import { existsSync, readFileSync } from "node:fs";
+import { existsSync, readFileSync } from "node:fs";
 import type { ExtensionAPI, ExtensionCommandContext } from "@earendil-works/pi-coding-agent";
 
 import { parseReviewArgs } from "./src/cli-args.js";

package/extensions/review/package.json

@@ -1,6 +1,6 @@
-{
+{
   "name": "@unifan/pi-review-zh",
-  "version": "1.0.0",
+  "version": "1.0.1",
   "description": "Pi AI 并发代码审查扩展（多专家子代理并发 + 门禁裁判系统，中文增强版）",
   "type": "module",
   "license": "MIT",
@@ -35,4 +35,4 @@
   "dependencies": {
     "@sinclair/typebox": "^0.34.0"
   }
-}
+}

package/extensions/review/src/report.ts

@@ -1,4 +1,4 @@
-import type {
+import type {
 	GateDisposition,
 	GateRunResult,
 	Issue,

package/extensions/review/src/tui-renderer.ts

@@ -1,4 +1,4 @@
-import { Box, Text } from "@earendil-works/pi-tui";
+import { Box, Text } from "@earendil-works/pi-tui";
 import type { ExtensionAPI } from "@earendil-works/pi-coding-agent";
 
 import type { IssueSeverity, Verdict } from "./types.js";

package/extensions/sessions/README.md

@@ -1,4 +1,4 @@
-# 📜 @unifan/pi-sessions-zh (历史会话管理器·中文版)
+# 📜 @unifan/pi-sessions-zh (历史会话管理器·中文版)
 
 为 **Pi Coding Agent** 打造的高颜值双栏历史会话选择器与实时预览恢复扩展。
 

package/extensions/sessions/index.ts

@@ -1,4 +1,4 @@
-import type { ExtensionAPI, ExtensionCommandContext } from "@earendil-works/pi-coding-agent";
+import type { ExtensionAPI, ExtensionCommandContext } from "@earendil-works/pi-coding-agent";
 import { DynamicBorder, SessionManager, keyHint } from "@earendil-works/pi-coding-agent";
 import {
   CancellableLoader,

package/extensions/sessions/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@unifan/pi-sessions-zh",
-  "version": "1.0.0",
+  "version": "1.0.1",
   "description": "Pi 历史会话管理器（双栏可视化 TUI 预览与恢复，中文增强版）",
   "type": "module",
   "main": "./index.ts",
@@ -18,4 +18,3 @@
   "author": "unifan",
   "license": "MIT"
 }
-

package/extensions/sessions/sessions.ts

@@ -1,4 +1,4 @@
-export interface SessionInfoLike {
+export interface SessionInfoLike {
   id: string;
   name?: string;
   cwd: string;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@unifan/pi-unifan-zh",
-  "version": "1.0.0",
+  "version": "1.0.2",
   "description": "Pi Coding Agent 中文扩展全家桶与独立插件集",
   "type": "module",
   "main": "extensions/sessions/index.ts",
@@ -33,4 +33,3 @@
   "author": "unifan",
   "license": "MIT"
 }
-

package/skills/README.md

@@ -1,4 +1,4 @@
-# 🎯 Pi Skills 目录 (自定义技能库)
+# 🎯 Pi Skills 目录 (自定义技能库)
 
 此目录用于存放你自定义的 Pi Agent Skills。