zen-gitsync 2.17.46 → 2.17.47

package/src/cli/ai/agent.js

@@ -1,1342 +1,1342 @@
-// Copyright 2026 xz333221
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
-//
-// g ai — 终端交互式 AI 编码智能体。
-//
-// 入口: runAiAgent(argv)(由 gitCommit.js 在 argv[2]==='ai' 时动态 import)。
-//
-// 能力:
-//   - 默认使用 g ui 里配置的模型(~/.zen-gitsync/config.json 顶层 models,
-//     isDefault 优先,否则取第一个);--model=<序号|名称> 或会话内 /model 切换
-//   - OpenAI 兼容流式 function calling:模型可以调用 run_command /
-//     read_file / write_file / edit_file / list_files / search_text / ask_user
-//   - 图片:Alt+V 粘贴剪贴板图片、/image <路径> 附加本地图片,
-//     以多模态(image_url)消息发给模型(需模型支持视觉)
-//   - 权限:启动目录内全开放,其他目录可读写;唯一红线由 safety.js 拦截
-//     (格式化磁盘、删根目录、关机等系统级破坏操作)
-//
-// 视觉风格(termui.js):对标 Codex CLI / Claude Code / OpenCode —
-//   盒式输入框、ora 等待 spinner、✻ 思考(橙黄斜体)、⚙ 工具头(青) + │ 结果槽、
-//   正文 ➤ 子弹头(绿)+ 逐行轻量 markdown(**bold** / `code` / # 标题 / ``` 围栏 / - 列表)
-//
-// 使用方式:
-//   g ai                  交互模式(REPL)
-//   g ai "帮我修个 bug"    单发模式:执行一轮后退出
-//   g ai --model=2        指定配置里的第 2 个模型
-//
-// 实现注意:
-//   - 本模块属于 CLI 侧,不依赖 GUI 服务器代码(src/ui/server/**)
-//   - Ctrl+C 执行中只取消当前任务；空闲时交给主 CLI 清理并退出
-
-import readline from 'node:readline'
-import chalk from 'chalk'
-import config from '../../config.js'
-import { getCwd } from '../../utils/index.js'
-import { registerCleanup } from '../cleanup.js'
-import { terminateCommand, TOOL_DEFINITIONS } from './tools.js'
-import { runAgentTurn } from './turn.js'
-import { AgentExtensions } from './extensions.js'
-import { repairToolHistory, loadProjectInstructions, resolveRequestBudget } from './context.js'
-import {
-  printBanner, printHelpPanel,
-  filterSlashCommands, renderSlashHintBody, parseKeyForSlashHint,
-  printTurnSummary, renderPasteEcho,
-  printOk, printWarn, printError, printDim,
-} from './termui.js'
-import { readClipboardImage, checkImageFile, formatBytes } from './images.js'
-import { installPasteCapture, enableBracketedPaste, disableBracketedPaste } from './paste.js'
-import { runModelSetup, collectModelInput, buildModelConfig, selectFromList } from './modelSetup.js'
-import { genSessionId, autoTitle, writeSession, enforceRetention, listSessions } from './sessionStore.js'
-
-// truncateDisplay 已迁移到 termui.js;这里 re-export 保持既有测试/外部引用不断
-export { truncateDisplay } from './termui.js'
-
-// ──────────────────────────────────────────────
-// i18n 字符串表(CLI 其他部分以中文为主,这里按 locale 给双语)
-// ──────────────────────────────────────────────
-const STRINGS = {
-  zh: {
-    noModel: '未配置 AI 模型,请先在 GUI 通用设置中添加模型(运行 g ui 打开设置)',
-    unknownModel: (q) => `未找到匹配的模型: ${q},使用默认模型`,
-    waiting: '思考中…',
-    llmError: (msg) => `LLM 请求失败: ${msg}`,
-    toolIterLimit: (n) => `已达单轮最大工具调用次数(${n}),本轮结束。如需继续请再发一条消息。`,
-    toolRunning: (name) => `执行 ${name}…`,
-    busy: '任务执行中，请稍候，或按 Ctrl+C 停止当前任务。',
-    bye: '已退出 g ai',
-    cleared: '对话历史已清空',
-    newConversation: '已开启新对话',
-    operationCancelled: '已取消当前操作',
-    thinkOn: '思考过程显示: 开',
-    thinkOff: '思考过程显示: 关(模型仍会思考,只是不回显)',
-    modelSwitched: (label) => `已切换模型: ${label}`,
-    invalidModelIndex: '无效的模型序号',
-    cdOk: (p) => `工作目录已切换: ${p}`,
-    cdFail: (p) => `目录不存在: ${p}`,
-    currentModel: '当前模型',
-    availableModels: '可用模型',
-    oneShotNoModel: '未配置模型,无法启动',
-    oneShotDone: '单发模式完成 · 需要多轮连续对话请直接运行 g ai(不带参数)',
-    helpTitle: 'g ai 命令',
-    bannerTip: 'Alt+V 贴图 · /help 帮助 · Ctrl+C 停止当前任务 · /exit 退出',
-    prompt: '❯ ',
-    bannerModel: '模型',
-    bannerCwd: '目录',
-    thinkingLabel: '思考',
-    answerLabel: '回答',
-    thinkingHint: '… 已显示前 12 行；输入 /think full 开启后续完整思考',
-    emptyResponse: '模型没有返回回答，请重试或切换模型。',
-    chars: '字符',
-    extLoadingMcp: (n) => `正在连接 ${n} 个 MCP 服务…`,
-    extLoaded: (skills, tools) => `扩展已就绪: ${skills} 个 Skill · ${tools} 个 MCP 工具(/skills 查看)`,
-    extNone: '',
-    extSkillTitle: '已安装的 Skill',
-    extSkillEmpty: '没有装 Skill。可以在 g ui → 智能体 → Skill 广场 里挑一个装到当前项目或 g ai。',
-    extMcpTitle: '已接入的 MCP 服务',
-    extMcpEmpty: '没有接入 MCP 服务。可以在 g ui → 智能体 → MCP 广场 里挑一个装到当前项目或 g ai。',
-    extMcpFailed: '连接失败',
-    imagePasting: '正在读取剪贴板图片…',
-    imageAttached: (n, size) => `📎 图片 #${n} 已附加(${size}),将随下一条消息发送`,
-    imageEmpty: '剪贴板中没有图片;可先截图再按 Alt+V,或用 /image <路径> 附加本地图片',
-    imageBadPath: (p) => `图片不存在或格式不支持: ${p}(支持 png/jpg/jpeg/gif/webp/bmp)`,
-    imageListTitle: '待发送图片',
-    imageListEmpty: '当前没有待发送的图片(Alt+V 粘贴或 /image <路径> 附加)',
-    imageCleared: '已清除待发送的图片',
-    imageSending: (n) => `📎 附带 ${n} 张图片`,
-    pasteToken: (n, lines) => `[粘贴 #${n} · ${lines} 行]`,
-    pasteExpired: (n) => `有 ${n} 段粘贴内容已从内存回收，无法还原；如仍需要请重新粘贴`,
-    addModelTitle: '添加模型配置',
-    addModelSaved: (name) => `✓ 模型 "${name}" 已添加并切换为当前模型`,
-    addModelCancelled: '已取消添加模型',
-    addModelSaveError: (msg) => `保存配置失败: ${msg}`,
-    resumeTitle: '选择要恢复的对话',
-    resumeEmpty: '没有可恢复的历史对话',
-    resumeCancelled: '已取消恢复对话',
-    resumeDone: (title, count) => `已恢复对话: ${title} (${count} 条消息)`,
-  },
-  en: {
-    noModel: 'No AI model configured. Add one in GUI settings first (run `g ui`).',
-    unknownModel: (q) => `No model matching "${q}", falling back to default`,
-    waiting: 'Thinking…',
-    llmError: (msg) => `LLM request failed: ${msg}`,
-    toolIterLimit: (n) => `Hit max tool iterations (${n}) for this turn. Send another message to continue.`,
-    toolRunning: (name) => `Running ${name}…`,
-    busy: 'Agent is working. Wait or press Ctrl+C to stop the task.',
-    bye: 'Bye',
-    cleared: 'Conversation cleared',
-    newConversation: 'Started a new conversation',
-    operationCancelled: 'Current operation cancelled',
-    thinkOn: 'Thinking display: on',
-    thinkOff: 'Thinking display: off (model still thinks, just hidden)',
-    modelSwitched: (label) => `Switched model: ${label}`,
-    invalidModelIndex: 'Invalid model index',
-    cdOk: (p) => `Working directory changed: ${p}`,
-    cdFail: (p) => `Directory not found: ${p}`,
-    currentModel: 'Current model',
-    availableModels: 'Available models',
-    oneShotNoModel: 'No model configured, aborting',
-    oneShotDone: 'One-shot done · for multi-turn chat run `g ai` with no arguments',
-    helpTitle: 'g ai commands',
-    bannerTip: 'Alt+V image · /help commands · Ctrl+C stop task · /exit quit',
-    prompt: '❯ ',
-    bannerModel: 'Model',
-    bannerCwd: 'CWD',
-    thinkingLabel: 'Thinking',
-    answerLabel: 'Answer',
-    thinkingHint: '… first 12 lines shown; enter /think full for future reasoning',
-    emptyResponse: 'The model returned no answer. Retry or switch models.',
-    chars: 'chars',
-    extLoadingMcp: (n) => `Connecting to ${n} MCP server(s)…`,
-    extLoaded: (skills, tools) => `Extensions ready: ${skills} skill(s) · ${tools} MCP tool(s) (/skills for details)`,
-    extNone: '',
-    extSkillTitle: 'Installed skills',
-    extSkillEmpty: 'No skills installed. Browse g ui → Agent → Skill marketplace to install one into this project or g ai.',
-    extMcpTitle: 'Connected MCP servers',
-    extMcpEmpty: 'No MCP servers connected. Browse g ui → Agent → MCP marketplace to install one.',
-    extMcpFailed: 'failed',
-    imagePasting: 'Reading clipboard image…',
-    imageAttached: (n, size) => `📎 Image #${n} attached (${size}); sent with your next message`,
-    imageEmpty: 'No image in clipboard; take a screenshot first, or use /image <path>',
-    imageBadPath: (p) => `Not a supported image: ${p} (png/jpg/jpeg/gif/webp/bmp)`,
-    imageListTitle: 'Pending images',
-    imageListEmpty: 'No pending images (Alt+V to paste, or /image <path>)',
-    imageCleared: 'Pending images cleared',
-    imageSending: (n) => `📎 ${n} image(s) attached`,
-    pasteToken: (n, lines) => `[paste #${n} · ${lines} lines]`,
-    pasteExpired: (n) => `${n} pasted block(s) were dropped from memory and could not be restored; paste again if still needed`,
-    addModelTitle: 'Add Model Configuration',
-    addModelSaved: (name) => `✓ Model "${name}" added and switched to`,
-    addModelCancelled: 'Add model cancelled',
-    addModelSaveError: (msg) => `Failed to save: ${msg}`,
-    resumeTitle: 'Choose a conversation to resume',
-    resumeEmpty: 'No previous conversations to resume',
-    resumeCancelled: 'Resume cancelled',
-    resumeDone: (title, count) => `Resumed: ${title} (${count} messages)`,
-  },
-}
-
-function makeStrings(locale) {
-  return String(locale || '').startsWith('en') ? STRINGS.en : STRINGS.zh
-}
-
-// 模型的展示名:优先 model 字段,其次 name,最后 baseURL
-function modelLabel(m) {
-  return m?.model || m?.name || m?.baseURL || '(unknown)'
-}
-
-// ──────────────────────────────────────────────
-// 系统 prompt — 明确告知环境、权限边界与工作方式
-// ──────────────────────────────────────────────
-// 顺序固定:基础人设 → 项目说明(AGENTS.md/CLAUDE.md)→ 扩展能力(skill 清单 + MCP 工具)。
-// 把扩展放最后,是因为它是"增量能力",不该插在权限与环境说明中间打断阅读。
-// state 里存着已加载的 AgentExtensions;没加载(如单测直接调)时 extra 为空,行为与改动前一致。
-function extensionSuffix(state) {
-  try {
-    return state?.extensions?.promptSuffix?.(state.locale) || ''
-  } catch {
-    return ''
-  }
-}
-
-async function buildProjectPrompt(options) {
-  return buildSystemPrompt(options) + await loadProjectInstructions(options.cwd) + (options.extra || '')
-}
-
-function buildSystemPrompt({ cwd, locale, shellDesc }) {
-  const zh = !String(locale || '').startsWith('en')
-  const now = new Date().toLocaleString()
-  const isWin = process.platform === 'win32'
-  // 工具数量动态取自 tools.js —— 之前这里写死过"6 个",加 ask_user 时没人想起来改,
-  // 就变成了对模型撒谎。数字跟着定义走,以后再加工具不会漏。
-  const builtinToolCount = TOOL_DEFINITIONS.length
-  if (zh) {
-    return `你是 "g ai" —— zen-gitsync CLI 内置的终端编码智能体,通过工具在用户真实电脑上完成编码任务。
-
-# 运行环境
-- 操作系统: ${process.platform}
-- Shell: ${shellDesc}
-- 当前工作目录: ${cwd}(用户在此启动 g ai,也是所有相对路径的基准)
-- 当前时间: ${now}
-
-# 平台兼容性(重要!)
-- 必须使用与当前 Shell 兼容的命令,禁止盲套 Unix 写法
-${isWin ? `- 当前是 Windows cmd.exe,以下 Unix 命令**不存在**,用了必定报"不是内部或外部命令":
-  tail / head / cat / grep / ls / sed / awk / wc / cut / uniq / xargs / which / touch
-  平台守卫会在执行前拦截这些命令,但请主动避免,不要浪费一轮调用
-- 跨平台替代方案:
-  · 列目录 → list_files 工具 或 cmd 的 dir
-  · 搜内容 → search_text 工具 或 cmd 的 findstr
-  · 看文件 → read_file 工具 或 cmd 的 type
-  · 看图片 → read_image 工具(截图/设计稿/示意图)。read_file 读图片只会得到乱码,别浪费那一轮
-  · 看输出末尾 → PowerShell "命令 | Select-Object -Last N"
-  · 文本处理 → node -e "..." 或 PowerShell
-  · 查命令路径 → cmd 的 where(不是 which)
-- 必须跑 shell 时优先跨平台写法(如 node -e "..."),别用 Unix 专属命令` : `- 当前是 POSIX 环境,Unix 命令可用`}
-
-# 远程仓库(GitHub / Gitee)
-- 问"这个项目关联哪个远端" → run_command 跑 \`git remote -v\`:本地信息,不联网、不依赖任何 CLI,任何环境都能答
-- 问"我账号下有哪些仓库"或某仓库的 PR / Issue → 用官方 CLI。凭据由 CLI 自己保管:
-  · GitHub → \`gh\`。列仓库 \`gh repo list --limit 50 --json name,visibility,updatedAt,primaryLanguage\`;
-    看详情 \`gh repo view <owner/repo>\`;PR \`gh pr list\`;登录态 \`gh auth status\`
-  · Gitee → \`gitee\`。列仓库 \`gitee repo list\`;登录态 \`gitee auth status\`
-    (未登录时退出码**仍是 0**,必须看 \`--json\` 里的 status 字段,只看退出码会误判成已登录)
-  · 两者默认只覆盖**当前登录账号**;组织仓库、私有仓库可能需要更大的 token scope,拉不到就如实说明,
-    不要用其他途径绕过
-  · 绝不向用户索要 token —— 也不要让用户把 token 粘进对话
-- CLI 可能没装、或没在当前进程的 PATH 里(装完没重启服务就是这个表现):
-  报"不是内部或外部命令" / command not found 时,**不要换写法反复重试**,更不要凭印象编仓库名或目录名。
-  直接告诉用户未检测到该 CLI,可在 ZenGitSync 的「远程仓库」页一键安装/登录;
-  若只是想回答本项目的问题,退回 \`git remote -v\`
-- 克隆仓库 / 添加远端时**优先用 SSH**(用户偏好,本机已配好密钥,实测走 https 会弹凭据窗口把任务打断):
-  · GitHub 用 \`git@github.com:owner/repo.git\`,Gitee 用 \`git@gitee.com:owner/repo.git\`
-  · 用户给的是 \`https://...\` 或 \`gh repo clone owner/repo\` 时,先换算成 SSH 地址再执行 ——
-    走 https 会弹 Git Credential Manager 让用户输账号密码,任务就停在半路等输入
-  · 已有仓库换协议:\`git remote set-url origin <ssh 地址>\`(先 \`git remote -v\` 看当前是什么)
-  · 只有 SSH 真的不可用(报 \`Permission denied (publickey)\` / \`Host key verification failed\`)才退回 https,
-    并用一句话说明这次走的是 https、配好密钥后可改回;两边都失败就停下来问用户,不要反复重试
-
-# 权限(用户已明确授权,无需反复征求同意)
-- 工作目录内:读写文件、执行命令等所有操作直接执行
-- 其他目录:同样可以读取和修改
-- 唯一红线:不得破坏系统(格式化磁盘、删除根目录/系统目录、关机重启、写块设备等)。
-  命令守卫只能识别部分危险写法，不是沙箱；不得通过脚本或嵌套 shell 绕过限制。
-
-# 工作方式
-- 先动手、后提问:能用工具查清的不要问用户(list_files / read_file / search_text / run_command)
-- 修改代码后主动验证:跑测试、构建或至少语法检查(用 run_command)
-- 编辑文件优先 edit_file 精确替换;先 read_file 看原文,old_string 必须与文件内容完全一致(含缩进换行)
-- 大文件用 offset/limit 分段读取,不要一次读爆上下文
-- git 操作用 run_command 执行;提交代码可以用 git 命令,也可以用本 CLI 的 g -y(默认信息提交并推送)或 g --ai(AI 生成提交信息)
-- run_command 默认就在工作目录执行,不要再加 cd / cd /d 前缀;默认超时 120 秒,长任务加大 timeout_seconds(最大 600)
-- 命令在 ${shellDesc} 下执行,注意语法兼容
-- 先读取当前项目的规范、依赖与测试配置，再选择对应的验证命令，不要假定所有项目都使用 npm
-- 用户要求概述或简短回答时，只读取必要的结构与入口文件，控制探索范围
-
-# 任务计划(多步任务必用)
-- 任务需要多步时(改代码、排查、调研、多文件改动),**动手之前**先调一次 update_plan,把要做的事
-  拆成 3-8 个可核对的步骤,让用户在你改任何东西之前就知道范围和验收标准
-- 单步小事(读个文件、答一个问题)不要列计划,那是噪音
-- 每次都传**完整**的当前计划(不是增量),顺序即执行顺序;同一时刻最多一个 in_progress
-- 完成一步就更新一次状态,别攒到最后一次性刷;计划与实际不符时直接改计划,不要硬着头皮往下走
-- 全部做完后把 steps 传空表示收尾
-- update_plan 只是给人看的进度板,不要为了"更新计划"去改文件或跑命令
-
-# 与用户交互
-- 需要向用户确认、提问或汇报重要决策时,直接用普通文本输出 —— 用户能实时看到你的文本;
-  需要暂停当前任务并等待用户决定或补充信息时,调用 ask_user,不要猜测或只在普通文本里提问
-  不要调用不存在的工具,内置工具就是上面列出的 ${builtinToolCount} 个(若用户装了 MCP,你的工具表里还会多出 mcp__ 开头的工具)
-- 图片有两条来路:
-  · 用户通过 Alt+V 或 /image 附加 → 图片以 image_url 部件出现在 user 消息里,直接就能看到
-  · 用户只给了**路径**(或你自己在翻文件时遇到图)→ 用 read_image 工具读它,别用 read_file
-    (read_file 按 UTF-8 解码,图片只会是乱码),也别根据文件名猜内容
-  如果当前模型不支持视觉(带图请求报错),提醒用户换用支持视觉的模型
-- 发现高风险或状态不一致的情况(例如版本号 / git tag / CHANGELOG 对不上、发布前环境异常、
-  仓库状态与预期不符)时:先用文本说明发现和影响,停下来等用户指示,不要擅自继续破坏性操作
-
-# 输出
-- 你的文本输出直接显示在用户终端,用简体中文交流
-- 完成任务后用一两句话汇报结果,不要复述过程细节`
-  }
-  return `You are "g ai" — the terminal coding agent built into the zen-gitsync CLI. You use tools to perform coding tasks on the user's real machine.
-
-# Environment
-- OS: ${process.platform}
-- Shell: ${shellDesc}
-- Working directory: ${cwd} (where the user launched g ai; base for all relative paths)
-- Current time: ${now}
-
-# Platform compatibility (important!)
-- You MUST use commands compatible with the current shell. Do NOT blindly copy Unix patterns.
-${isWin ? `- This is Windows cmd.exe. The following Unix commands do NOT exist here and will fail with "not recognized":
-  tail / head / cat / grep / ls / sed / awk / wc / cut / uniq / xargs / which / touch
-  A platform guard will block these before execution, but avoid them proactively — don't waste a turn.
-- Cross-platform alternatives:
-  · List dirs → list_files tool, or cmd's dir
-  · Search content → search_text tool, or cmd's findstr
-  · Read files → read_file tool, or cmd's type
-  · Look at an image → read_image tool (screenshots, mockups, diagrams). read_file only decodes text and returns garbage for images
-  · Tail output → PowerShell "command | Select-Object -Last N"
-  · Text processing → node -e "..." or PowerShell
-  · Find executable → cmd's where (not which)
-- For shell one-offs, prefer cross-platform forms like node -e "..."` : `- POSIX environment: Unix commands are available`}
-
-# Remote repositories (GitHub / Gitee)
-- "Which remote does this project point at?" → run_command \`git remote -v\`: local info, no network, no CLI needed
-- "Which repos do I have?" or a repo's PRs / issues → use the official CLI. The CLI owns the credentials:
-  · GitHub → \`gh\`. List repos \`gh repo list --limit 50 --json name,visibility,updatedAt,primaryLanguage\`;
-    details \`gh repo view <owner/repo>\`; PRs \`gh pr list\`; auth state \`gh auth status\`
-  · Gitee → \`gitee\`. List repos \`gitee repo list\`; auth state \`gitee auth status\`
-    (the exit code is still 0 when logged out — read the status field from \`--json\`, or you will
-    wrongly report "logged in")
-  · Both default to the **currently authenticated account only**; org and private repos may need
-    broader token scopes. If you cannot see them, say so plainly — do not route around it
-  · Never ask the user for a token, and never have them paste one into the conversation
-- The CLI may not be installed, or missing from this process's PATH (that is what "installed but the
-  server cannot find it" looks like):
-  on "not recognized" / command not found, do NOT retry with a different spelling and do NOT invent
-  repo or directory names. Tell the user the CLI was not found and point them to ZenGitSync's
-  "Remote repositories" page (one-click install / login); if you only need to answer for this project,
-  fall back to \`git remote -v\`
-- Cloning a repo or adding a remote: **prefer SSH** (the user's preference — keys are already set up,
-  and https has been observed to pop a credential window that stalls the task):
-  · GitHub → \`git@github.com:owner/repo.git\`; Gitee → \`git@gitee.com:owner/repo.git\`
-  · If the user hands you an \`https://...\` URL (or \`gh repo clone owner/repo\`), convert it to the SSH
-    form first — https pops up Git Credential Manager asking for a username/password and blocks the task
-  · Switching an existing repo: \`git remote set-url origin <ssh url>\` (check \`git remote -v\` first)
-  · Fall back to https only when SSH genuinely fails (\`Permission denied (publickey)\` /
-    \`Host key verification failed\`), and say in one line that this one used https and can go back to SSH
-    once the key is set up; if both fail, stop and ask the user — do not keep retrying
-
-# Permissions (explicitly granted by the user — do not keep asking)
-- Inside the working directory: read/write files and run commands directly
-- Other directories: may also be read and modified
-- Never destroy the system (format disks, delete root/system dirs, shutdown/reboot, write block devices). The command guard is not a sandbox; never bypass it with scripts or nested shells.
-
-# How to work
-- Act first, ask later: use tools (list_files / read_file / search_text / run_command) instead of asking the user
-- When a decision or missing detail must come from the user, call ask_user. Do not guess or merely describe a question in plain text.
-- After editing code, verify: run tests, builds, or at least a syntax check via run_command
-- Prefer edit_file for precise replacements; read_file first, old_string must match the file exactly
-- Read large files in segments (offset/limit)
-- Git operations go through run_command; to commit, use git commands or this CLI's g -y / g --ai
-- run_command already executes in the working directory — do NOT prefix with cd; default timeout 120s, raise timeout_seconds (max 600) for long tasks
-- Commands run under ${shellDesc}; keep syntax compatible
-- Read this project's instructions, dependencies and test configuration before choosing verification commands; do not assume npm
-- For a brief overview, inspect only the necessary structure and entry points; keep exploration proportional to the request
-
-# Task plan (required for multi-step work)
-- When a task takes several steps (code changes, debugging, research, multi-file edits), call update_plan
-  **before touching anything**: break it into 3-8 verifiable steps so the user knows the scope and the
-  acceptance criteria before you modify a single file
-- Do not plan single trivial actions (read one file, answer one question) — that is noise
-- Always send the **complete** current plan (not a delta), in execution order; at most one in_progress
-- Update the status right after finishing a step instead of batching everything at the end; when the
-  plan no longer matches reality, rewrite the plan instead of pushing ahead
-- Send an empty steps list once everything is done
-- update_plan is a progress board for humans: never edit files or run commands just to "update the plan"
-
-# Talking to the user
-- When you need to confirm something, ask a question, or report an important decision, just write plain text — the user sees your output in real time. Use ask_user when the task must pause for an answer. Never call tools that do not exist; the built-in set is the ${builtinToolCount} tools listed above (if the user installed MCP servers, extra tools prefixed with mcp__ will also appear in your tool list)
-- Images reach you two ways:
-  · The user attached one via Alt+V or /image → it arrives as an image_url part in the user message; you can see it directly
-  · The user only gave you a **path** (or you run into an image while exploring) → read it with the read_image tool, not read_file
-    (read_file decodes UTF-8 and returns garbage for images), and never guess an image's content from its filename
-  If the current model rejects images (no vision support), tell the user to switch to a vision-capable model
-- When you spot high-risk or inconsistent state (version number / git tag / CHANGELOG mismatch, abnormal release environment, unexpected repo state), explain the finding and its impact in text, then STOP and wait for the user's decision instead of proceeding with destructive operations
-
-# Output
-- Your text output goes straight to the user's terminal; reply in English
-- Report results in one or two sentences; do not narrate the process`
-}
-
-// ──────────────────────────────────────────────
-// 消息准备(消毒 / 旧图片降级)统一实现在 context.js —— CLI 与 Web 智能体面板
-// 共用同一份,避免两侧各写一遍后行为分叉。这里 re-export 保持既有引用与测试不断。
-// ──────────────────────────────────────────────
-export { sanitizeMessages, stripStaleImages } from './context.js'
-
-// ──────────────────────────────────────────────
-// 单轮 agent 循环:用户一句话 → 流式输出 → 工具调用 → 再调用模型 … 直到模型给出最终文本
-// images: [{path, bytes}] 待发送图片(可为空数组)
-// ──────────────────────────────────────────────
-function printSlashHelp(t, locale) {
-  const zh = !String(locale || '').startsWith('en')
-  const lines = zh ? [
-    '  /help             显示本帮助',
-    '  /model            列出可用模型;/model <序号> 切换模型',
-    '  /addmodel         添加新的模型配置(交互式向导)',
-    '  /cd <路径>        切换智能体工作目录',
-    '  /image [路径]     查看待发送图片;/image <路径> 附加本地图片;/image clear 清除',
-    '  /think full       完整显示后续思考（默认，模型返回的内容）',
-    '  /think compact    预览前 12 行思考',
-    '  /think off        隐藏思考；/think 切换完整与隐藏',
-    '  /tools [模式]     compact 精简 / full 完整工具输出',
-    '  /skills           查看已安装的 Skill 与已接入的 MCP 服务',
-    '  /stats            查看上一轮与会话累计用量、耗时',
-    '  /new              开启新对话',
-    '  /resume           选择并恢复之前的对话',
-    '  /clear            清空对话历史',
-    '  /exit, /quit      退出',
-    '',
-    '  Alt+V             粘贴剪贴板图片,随下一条消息发送(需视觉模型)',
-    '  多行粘贴          直接粘贴多行文本:整段作为一条消息发送,输入行里显示 [粘贴 #N] 占位符',
-    '',
-    '权限说明: 工作目录内全部操作直接执行;其他目录可读写;',
-    '命令守卫识别部分系统级危险操作，不是隔离沙箱。',
-  ] : [
-    '  /help             Show this help',
-    '  /model            List models; /model <n> to switch',
-    '  /addmodel         Add a new model (interactive wizard)',
-    '  /cd <path>        Change agent working directory',
-    '  /image [path]     List pending images; attach a file; /image clear to reset',
-    '  /think full       Show future thinking returned by the model in full (default)',
-    '  /think compact    Preview the first 12 thinking lines',
-    '  /think off        Hide thinking; /think toggles full / hidden',
-    '  /tools [mode]     compact / full tool output',
-    '  /skills           List installed skills and connected MCP servers',
-    '  /stats            Last turn timing and session token usage',
-    '  /new              Start a new conversation',
-    '  /resume           Choose and resume a previous conversation',
-    '  /clear            Clear conversation history',
-    '  /exit, /quit      Quit',
-    '',
-    '  Alt+V             Paste clipboard image (needs a vision-capable model)',
-    '  Multi-line paste  Paste a block as one message; the input line shows a [paste #N] placeholder',
-    '',
-    'Permissions: full access in the working directory; other dirs readable/writable;',
-    'the command guard detects some dangerous operations; it is not a sandbox.',
-  ]
-  printHelpPanel(t.helpTitle, lines)
-}
-
-export async function handleSlashCommand(state, input, t) {
-  const [cmd, ...rest] = input.split(/\s+/)
-  const arg = rest.join(' ').trim()
-  const zh = !String(state.locale).startsWith('en')
-  if (state.busy && !['/help', '/stats', '/think', '/tools', '/skills', '/mcp', '/exit', '/quit'].includes(cmd)) {
-    printDim(t.busy)
-    return 'ok'
-  }
-
-  if (cmd === '/exit' || cmd === '/quit') return 'exit'
-  if (cmd === '/help') { printSlashHelp(t, state.locale); return 'ok' }
-  if (cmd === '/clear' || cmd === '/new') {
-    await state.persistSession?.()
-    state.messages = [state.messages[0]]
-    // 开新会话 ID,旧会话保留在磁盘上
-    state.sessionId = genSessionId()
-    state.sessionCreatedAt = new Date().toISOString()
-    state.lastTurnStats = null
-    state.sessionStats = null
-    printOk(cmd === '/new' ? t.newConversation : t.cleared)
-    return 'ok'
-  }
-  if (cmd === '/resume') {
-    if (state.busy) {
-      printDim(t.busy)
-      return 'ok'
-    }
-    const sessions = (await listSessions())
-      .filter(session => session.sessionId !== state.sessionId && session.messages.length > 1)
-    if (sessions.length === 0) {
-      printDim(t.resumeEmpty)
-      return 'ok'
-    }
-
-    const zh = !String(state.locale || '').startsWith('en')
-    const cleanLabel = (value, fallback) => String(value || fallback).replace(/[\r\n\x00-\x1f\x7f]/g, ' ').trim()
-    const items = sessions.map((session) => {
-      const title = cleanLabel(session.title, zh ? '(未命名对话)' : '(untitled)')
-      const updated = new Date(session.updatedAt || session.createdAt || 0)
-      const date = Number.isNaN(updated.getTime()) ? '' : updated.toLocaleString(state.locale)
-      const count = session.messages.filter(message => message?.role !== 'system').length
-      return {
-        label: `${title}  ${chalk.dim(`${date} · ${count} ${zh ? '条消息' : 'messages'}`)}`,
-        value: session,
-      }
-    })
-    const selectStrings = {
-      navHint: zh ? '↑↓ 切换,Enter 确认' : '↑↓ navigate, Enter confirm',
-      selectPrompt: (min, max) => zh ? `请选择(${min}-${max}):` : `Choose (${min}-${max}):`,
-      invalidChoice: (min, max) => zh ? `请输入 ${min}-${max} 之间的序号` : `Enter a number from ${min} to ${max}`,
-    }
-    const asker = {
-      ask(prompt) {
-        return new Promise(resolve => state.rl.question(`${prompt} `, answer => resolve(String(answer || '').trim())))
-      },
-    }
-
-    state.inWizard = true
-    try {
-      const selected = await selectFromList({
-        rl: state.rl,
-        asker,
-        title: t.resumeTitle,
-        items,
-        t: selectStrings,
-      })
-      if (!selected?.value) return 'ok'
-
-      const session = selected.value
-      const { promises: fsp } = await import('node:fs')
-      if (session.cwd) {
-        try {
-          const stat = await fsp.stat(session.cwd)
-          if (stat.isDirectory()) state.ctx.cwd = session.cwd
-        } catch { /* 已删除的旧目录不阻断恢复,沿用当前目录 */ }
-      }
-      state.sessionId = session.sessionId
-      state.sessionCreatedAt = session.createdAt || new Date().toISOString()
-      state.messages = repairToolHistory(session.messages)
-      state.lastTurnStats = session.lastTurnStats || null
-      state.sessionStats = session.sessionStats || null
-      const systemMessage = {
-        role: 'system',
-        content: await buildProjectPrompt({ cwd: state.ctx.cwd, locale: state.locale, shellDesc: state.shellDesc, extra: extensionSuffix(state) }),
-      }
-      if (state.messages[0]?.role === 'system') state.messages[0] = systemMessage
-      else state.messages.unshift(systemMessage)
-      state.pendingImages = []
-      const count = state.messages.filter(message => message?.role !== 'system').length
-      printOk(t.resumeDone(cleanLabel(session.title, session.sessionId), count))
-    } catch (err) {
-      if (err?.cancelled) printDim(t.resumeCancelled)
-      else printError(err.message)
-    } finally {
-      state.inWizard = false
-    }
-    return 'ok'
-  }
-  if (cmd === '/think') {
-    const next = arg || (state.showThinking ? 'off' : 'full')
-    if (!['off', 'compact', 'full'].includes(next)) {
-      printWarn('/think compact | full | off')
-      return 'ok'
-    }
-    state.showThinking = next !== 'off'
-    state.thinkingMode = next
-    printOk(zh ? `后续请求的思考显示：${({ off: '隐藏', compact: '预览前 12 行', full: '完整' })[next]}` : `Thinking for future requests: ${next}`)
-    return 'ok'
-  }
-  if (cmd === '/tools') {
-    if (arg && !['full', 'compact'].includes(arg)) { printWarn('/tools compact | full'); return 'ok' }
-    state.fullTools = arg ? arg === 'full' : !state.fullTools
-    printOk(zh ? `后续工具结果：${state.fullTools ? '完整' : '精简'}` : `Future tool output: ${state.fullTools ? 'full' : 'compact'}`)
-    return 'ok'
-  }
-  if (cmd === '/skills' || cmd === '/mcp') {
-    const info = state.extensions?.summary(state.locale) || { skills: [], servers: [] }
-    console.log(chalk.bold(`\n  ${t.extSkillTitle}`))
-    if (info.skills.length === 0) printDim(t.extSkillEmpty)
-    for (const skill of info.skills) {
-      const scope = skill.scope === 'project' ? (zh ? '项目' : 'project') : (zh ? '全局' : 'global')
-      console.log(`  · ${chalk.cyan(skill.name)} ${chalk.dim(`[${scope}]`)}${skill.description ? chalk.dim(` — ${skill.description}`) : ''}`)
-    }
-    console.log(chalk.bold(`\n  ${t.extMcpTitle}`))
-    if (info.servers.length === 0) printDim(t.extMcpEmpty)
-    for (const server of info.servers) {
-      if (server.error) {
-        console.log(`  · ${chalk.red(server.id)} ${chalk.dim(`— ${t.extMcpFailed}: ${server.error}`)}`)
-      } else {
-        console.log(`  · ${chalk.cyan(server.id)} ${chalk.dim(`— ${server.tools} ${zh ? '个工具' : 'tools'} (${server.command})`)}`)
-      }
-    }
-    console.log('')
-    return 'ok'
-  }
-  if (cmd === '/stats') {
-    if (state.lastTurnStats) printTurnSummary(state.lastTurnStats, { locale: state.locale, session: state.sessionStats })
-    else printDim(zh ? '完成一轮对话后显示用量与耗时。' : 'Usage and timing appear after a turn completes.')
-    return 'ok'
-  }
-  if (cmd === '/image') {
-    if (arg === 'clear') {
-      state.pendingImages = []
-      printOk(t.imageCleared)
-      return 'ok'
-    }
-    if (!arg) {
-      if (state.pendingImages.length === 0) {
-        printDim(t.imageListEmpty)
-      } else {
-        console.log(chalk.bold(t.imageListTitle) + ':')
-        state.pendingImages.forEach((img, i) => {
-          console.log(`  ${i + 1}. ${img.path} ${chalk.dim('(' + formatBytes(img.bytes) + ')')}`)
-        })
-      }
-      return 'ok'
-    }
-    const { resolve } = await import('node:path')
-    const target = resolve(state.ctx.cwd, arg)
-    const img = await checkImageFile(target)
-    if (!img) {
-      printError(t.imageBadPath(target))
-      return 'ok'
-    }
-    state.pendingImages.push(img)
-    printOk(t.imageAttached(state.pendingImages.length, formatBytes(img.bytes)))
-    return 'ok'
-  }
-  if (cmd === '/model') {
-    if (!arg) {
-      console.log(chalk.bold(t.availableModels) + ':')
-      state.models.forEach((m, i) => {
-        const current = m === state.model ? chalk.green(' ← ' + t.currentModel) : ''
-        const keyHint = m.apiKey ? '' : chalk.yellow(' (无 apiKey)')
-        console.log(`  ${i + 1}. ${modelLabel(m)} ${chalk.dim(m.baseURL || '')}${keyHint}${current}`)
-      })
-      console.log(chalk.dim('/model <序号> 切换'))
-      return 'ok'
-    }
-    const idx = Number.parseInt(arg, 10)
-    if (!Number.isFinite(idx) || idx < 1 || idx > state.models.length) {
-      printError(t.invalidModelIndex)
-      return 'ok'
-    }
-    state.model = state.models[idx - 1]
-    printOk(t.modelSwitched(modelLabel(state.model)))
-    return 'ok'
-  }
-  if (cmd === '/addmodel') {
-    // 智能体执行中不允许启动向导:向导的 rl.question 会与流式输出交错,体验混乱
-    if (state.busy) {
-      printDim(t.busy)
-      return 'ok'
-    }
-    state.inWizard = true
-    try {
-      console.log(chalk.cyan.bold('\n' + t.addModelTitle))
-      // 复用 modelSetup 的交互式收集逻辑(服务商 → 接口 → 模型名 → Key → 显示名 → 测试)
-      const collected = await collectModelInput({
-        locale: state.locale,
-        rl: state.rl,
-        cancelMessage: t.addModelCancelled,
-      })
-      if (!collected) return 'ok'
-      const { baseURL, model, apiKey, displayName } = collected
-      // 构建模型配置(isDefault=false:/addmodel 只是追加,不改默认模型)
-      const newModel = buildModelConfig({ baseURL, model, apiKey, name: displayName, isDefault: false })
-      // 持久化到 ~/.zen-gitsync/config.json 顶层 models 数组
-      const cfg = await config.loadConfig()
-      const models = Array.isArray(cfg.models) ? cfg.models : []
-      models.push(newModel)
-      cfg.models = models
-      try {
-        await config.saveConfig(cfg)
-      } catch (err) {
-        printError(t.addModelSaveError(err.message))
-        return 'ok'
-      }
-      // 同步 REPL 状态:更新模型列表 + 自动切换到新模型(用户刚添加,大概率想立刻用)
-      state.models = models
-      state.model = newModel
-      printOk(t.addModelSaved(displayName || model))
-    } catch (err) {
-      printError(err.message)
-    } finally {
-      state.inWizard = false
-    }
-    return 'ok'
-  }
-  if (cmd === '/cd') {
-    if (!arg) { console.log(state.ctx.cwd); return 'ok' }
-    const { resolve } = await import('node:path')
-    const { promises: fsp } = await import('node:fs')
-    const target = resolve(state.ctx.cwd, arg)
-    try {
-      const st = await fsp.stat(target)
-      if (!st.isDirectory()) throw new Error('not a dir')
-      state.ctx.cwd = target
-      // 项目级 skill 与项目级 mcp.json 跟着工作目录走:先重载扩展(会关掉旧目录起的 MCP 子进程),
-      // 再重建首条 system 消息 —— 顺序反了会把旧目录的 skill 清单写进新目录的提示词。
-      await state.extensions?.reload({ cwd: target, locale: state.locale, onWarn: printWarn })
-      state.messages[0] = { role: 'system', content: await buildProjectPrompt({ cwd: target, locale: state.locale, shellDesc: state.shellDesc, extra: extensionSuffix(state) }) }
-      await state.persistSession?.()
-      printOk(t.cdOk(target))
-    } catch {
-      printError(t.cdFail(target))
-    }
-    return 'ok'
-  }
-  printWarn(`未知命令: ${cmd},输入 /help 查看可用命令`)
-  return 'ok'
-}
-
-// ──────────────────────────────────────────────
-// 主入口
-// ──────────────────────────────────────────────
-export async function runAiAgent(argv = []) {
-  const cfg = await config.loadConfig()
-  const locale = cfg.locale || 'zh-CN'
-  const t = makeStrings(locale)
-
-  // g ai --help:智能体专属帮助,不触发主 CLI 的 showHelp
-  if (argv.includes('--help') || argv.includes('-h')) {
-    printSlashHelp(t, locale)
-    return
-  }
-
-  let models = Array.isArray(cfg.models) ? cfg.models : []
-  if (models.length === 0) {
-    // 未配置模型:启动交互式配置向导,参照 g ui 设置里的"添加模型"
-    const setupResult = await runModelSetup({ locale })
-    if (!setupResult) {
-      process.exitCode = 1
-      return
-    }
-    models = setupResult.models
-  }
-
-  // --model=<序号|名称> 选择模型;默认 isDefault,否则第一个
-  let model = models.find(m => m.isDefault) || models[0]
-  const modelArg = argv.find(a => a.startsWith('--model='))
-  if (modelArg) {
-    const q = modelArg.split('=')[1]
-    const idx = Number.parseInt(q, 10)
-    if (Number.isFinite(idx) && idx >= 1 && idx <= models.length) {
-      model = models[idx - 1]
-    } else {
-      const found = models.find(m => modelLabel(m).toLowerCase().includes(String(q).toLowerCase()))
-      if (found) model = found
-      else printWarn(t.unknownModel(q))
-    }
-  }
-
-  const cwd = getCwd()
-  const shellDesc = process.platform === 'win32'
-    ? 'cmd.exe (Windows CMD)'
-    : (process.env.SHELL || '/bin/sh')
-
-  // 加载扩展(Skill 清单 + MCP 服务)。MCP 要起子进程,npx 首次还得下载包,
-  // 所以这一步可能明显停顿 —— 有 MCP 时先提示一句。任何失败都只降级并打印原因。
-  const extensions = await AgentExtensions.load({
-    cwd,
-    locale,
-    onWarn: printWarn,
-    onNotice: (count) => printDim(t.extLoadingMcp(count)),
-  })
-
-  const state = {
-    // 会话持久化:CLI 对话也保存到 ~/.zen-gitsync/agent-sessions/,
-    // 在 Web UI 智能体 tab 中可见(带 CLI 标记)
-    sessionId: genSessionId(),
-    sessionCreatedAt: new Date().toISOString(),
-    messages: [{ role: 'system', content: await buildProjectPrompt({ cwd, locale, shellDesc, extra: extensions.promptSuffix(locale) }) }],
-    ctx: {
-      cwd,
-      onChild: (child) => {
-        state.currentChild = child
-        // 子进程退出后清掉引用,避免 SIGINT 时 kill 一个已结束的进程
-        child.once?.('exit', () => { if (state.currentChild === child) state.currentChild = null })
-      },
-    },
-    models,
-    model,
-    locale,
-    shellDesc,
-    // 单轮工具调用上限:来自全局配置 aiMaxToolIterations(loadConfig 已规范化),
-    // 读不到时由 runAgentTurn 兜底为 200
-    maxToolIterations: cfg.aiMaxToolIterations,
-    // 单轮请求的上下文预算(字符上限/条数上限/单条 user 截断线)。解析公式与
-    // 默认值都在 context.js 的 resolveRequestBudget —— CLI 与 Web 共用同一份,
-    // 别在这里另算一套。turn.js 每轮把它透传给 prepareRequestMessages。
-    requestBudget: resolveRequestBudget(cfg.aiMaxRequestChars),
-    currentChild: null,
-    abortController: null,
-    cancelRequested: false,
-    busy: false,
-    showThinking: true,     // /think 切换:是否回显模型的思考过程
-    thinkingMode: 'full',
-    fullTools: false,
-    lastTurnStats: null,
-    sessionStats: null,
-    pendingImages: [],      // Alt+V / /image 附加的待发送图片 [{path, bytes}]
-    pasting: false,         // 剪贴板读取进行中(防 Alt+V 连打并发)
-    inWizard: false,        // /addmodel 交互式向导进行中:忽略 REPL 的 line 事件
-    rl: null,               // REPL readline 引用(供 /addmodel 复用做 rl.question)
-    extensions,             // 已加载的 Skill + MCP 扩展(见 extensions.js)
-  }
-
-  // 持久化当前会话到磁盘(供 Web UI 读取)
-  let saving = Promise.resolve()
-  async function persistSession() {
-    if (state.messages.length <= 1) return
-    try {
-      const title = autoTitle(state.messages)
-      const sessionId = state.sessionId
-      const snapshot = JSON.parse(JSON.stringify({
-        version: 1,
-        sessionId: state.sessionId,
-        title,
-        source: 'cli',
-        cwd: state.ctx.cwd,
-        model: modelLabel(state.model),
-        createdAt: state.sessionCreatedAt,
-        updatedAt: new Date().toISOString(),
-        messages: repairToolHistory(state.messages),
-        lastTurnStats: state.lastTurnStats,
-        sessionStats: state.sessionStats,
-      }))
-      saving = saving.catch(() => {}).then(() => writeSession(sessionId, snapshot))
-      await saving
-      enforceRetention().catch(() => {})
-    } catch (err) {
-      // 持久化失败不影响 CLI 正常使用
-      printDim(`(会话保存失败: ${err.message})`)
-    }
-  }
-  state.persistSession = persistSession
-
-  // SIGINT 时中止进行中的 LLM 请求 + 正在跑的子命令
-  // (进程退出与全局子进程 drain 由 gitCommit.js 的统一 handler 负责,这里只做快速止血)
-  registerCleanup('aiAgent', async () => {
-    state.cancelRequested = true
-    try { state.abortController?.abort() } catch (_) {}
-    await terminateCommand(state.currentChild)
-    // Ctrl+C 硬退时终端仍停在 bracketed paste 模式 → 之后 shell 里粘贴会带出 200~/201~ 乱码。
-    // 这个开关是终端层的,进程死了没人替它关,必须在这里补一刀。
-    disableBracketedPaste()
-    // MCP 子进程必须显式收掉:它们挂在 stdio 上,不收会让进程无法自然退出
-    await extensions.close()
-    await persistSession()
-  })
-
-  // 扩展摘要:一条 dim 行,没有装扩展时完全静默(extNone 为空串)
-  const extSummary = extensions.skills.length || extensions.tools.length
-    ? t.extLoaded(extensions.skills.length, extensions.tools.length)
-    : t.extNone
-
-  // 单发模式:g ai "帮我改个 bug"(位置参数拼成 prompt,跑一轮就退出)
-  const oneShot = argv.filter(a => !a.startsWith('-')).join(' ').trim()
-  if (oneShot) {
-    console.log(chalk.dim(`[${t.bannerModel}] ${modelLabel(model)} · [${t.bannerCwd}] ${cwd}`))
-    if (extSummary) printDim(extSummary)
-    state.abortController = new AbortController()
-    const stats = await runAgentTurn(state, oneShot, t)
-    if (stats.status === 'completed') printDim(t.oneShotDone)
-    else process.exitCode = stats.status === 'cancelled' ? 130 : 1
-    return
-  }
-
-  // ── 交互模式(REPL)──
-  printBanner({
-    title: 'Zen GitSync — AI Agent',
-    modelLabel: modelLabel(model),
-    baseURL: model.baseURL || '',
-    cwd,
-    modelText: t.bannerModel,
-    cwdText: t.bannerCwd,
-    tip: t.bannerTip,
-  })
-  if (extSummary) printDim(extSummary)
-
-  const rl = readline.createInterface({
-    input: process.stdin,
-    output: process.stdout,
-    // 颜色闭合，避免提示符的样式泄漏到后续输出。
-    prompt: chalk.cyan.bold(t.prompt + ' '),
-    historySize: 200,
-  })
-  state.rl = rl  // 供 /addmodel 等需要 rl.question 的斜杠命令复用
-
-  // 多行粘贴:终端里粘一整段文本时,readline 会把每个 \n 当 Enter —— 一次粘贴变成多次提交,
-  // 第 2 条起还会撞上下面的 busy 分支被丢掉。这里在数据进 readline 之前把整段捕获成占位符,
-  // 提交时再还原(详见 paste.js)。必须在处理 line 事件之前装好。
-  const pasteCapture = installPasteCapture(rl, {
-    formatToken: ({ index, lines }) => t.pasteToken(index, lines),
-  })
-
-  // ask_user pauses the tool loop inside the same readline session. The REPL
-  // line handler is bypassed while this small wizard owns the input line.
-  state.ctx.askUser = ({ question, options = [], allowFreeText = true, multiple = false }) => {
-    const zh = !String(state.locale || '').startsWith('en')
-    const safeOptions = Array.isArray(options) ? options.filter(Boolean).slice(0, 20) : []
-    // 多选只在真的给了选项时才有意义（与 tools.js 的归一化口径一致）
-    const multi = multiple === true && safeOptions.length > 0
-    state.inWizard = true
-    if (safeOptions.length > 0) {
-      console.log(chalk.cyan(`\n${question}`))
-      safeOptions.forEach((option, index) => console.log(`  ${chalk.cyan(String(index + 1))}. ${option}`))
-    } else {
-      console.log(chalk.cyan(`\n${question}`))
-    }
-
-    return new Promise(resolve => {
-      let settled = false
-      const questionController = new AbortController()
-      const parentSignal = state.abortController?.signal
-      const cleanup = () => {
-        rl.removeListener?.('close', onClose)
-        rl.removeListener?.('operationCancel', onCancel)
-        rl.input?.removeListener?.('keypress', onKeypressCancel)
-        parentSignal?.removeEventListener?.('abort', onParentAbort)
-      }
-      const finish = answer => {
-        if (settled) return
-        settled = true
-        cleanup()
-        state.inWizard = false
-        resolve(answer)
-      }
-      const cancel = () => {
-        if (settled) return
-        questionController.abort()
-        finish('Cancelled by user; the question was not answered.')
-      }
-      const onClose = () => cancel()
-      const onCancel = () => {
-        state.cancelRequested = true
-        try { state.abortController?.abort() } catch (_) {}
-        cancel()
-      }
-      const onParentAbort = () => cancel()
-      const onKeypressCancel = (_str, key) => {
-        if (!key?.ctrl || key.name !== 'c') return
-        key.name = 'escape'
-        key.sequence = ''
-        key.ctrl = false
-        key.meta = false
-        key.shift = false
-        onCancel()
-      }
-
-      rl.once('close', onClose)
-      rl.once('operationCancel', onCancel)
-      rl.input?.prependListener?.('keypress', onKeypressCancel)
-      parentSignal?.addEventListener?.('abort', onParentAbort, { once: true })
-
-      const promptAnswer = () => {
-        const hint = safeOptions.length > 0
-          ? (multi
-              ? (zh ? '请输入序号(逗号分隔可多选)' : 'Choose numbers, comma-separated for multiple')
-              : allowFreeText ? (zh ? '请输入序号或直接输入回答' : 'Choose a number or type an answer') : (zh ? '请输入序号' : 'Choose a number'))
-          : (zh ? '请输入回答' : 'Type your answer')
-        rl.question(`${hint}: `, { signal: questionController.signal }, answer => {
-          if (settled) return
-          const trimmed = String(answer || '').trim()
-          if (multi) {
-            // 逗号分隔的序号 → 选项列表(JSON 数组字符串回给模型,与 Web 端口径一致);
-            // 整行不是合法序号串时,允许自由输入的话就把它当作唯一一项
-            const parts = trimmed.split(/[,，]/).map(part => part.trim()).filter(Boolean)
-            const indexes = parts.map(part => Number.parseInt(part, 10))
-            const allIndexes = parts.length > 0 && indexes.every((n, i) => Number.isInteger(n) && String(n) === parts[i] && n >= 1 && n <= safeOptions.length)
-            if (allIndexes) {
-              finish(JSON.stringify(indexes.map(n => safeOptions[n - 1])))
-              return
-            }
-            if (allowFreeText && trimmed) {
-              finish(JSON.stringify([trimmed]))
-              return
-            }
-            console.log(chalk.yellow(zh ? `请输入 1-${safeOptions.length} 之间的序号,多个用逗号分隔` : `Enter numbers from 1 to ${safeOptions.length}, comma-separated`))
-            promptAnswer()
-            return
-          }
-          if (safeOptions.length > 0) {
-            const index = Number.parseInt(trimmed, 10)
-            if (Number.isInteger(index) && index >= 1 && index <= safeOptions.length) {
-              finish(safeOptions[index - 1])
-              return
-            }
-            if (!allowFreeText || !trimmed) {
-              console.log(chalk.yellow(zh ? `请输入 1-${safeOptions.length} 之间的序号` : `Enter a number from 1 to ${safeOptions.length}`))
-              promptAnswer()
-              return
-            }
-          }
-          if (!trimmed) {
-            console.log(chalk.yellow(zh ? '回答不能为空' : 'Answer cannot be empty'))
-            promptAnswer()
-            return
-          }
-          finish(trimmed)
-        })
-      }
-      promptAnswer()
-    })
-  }
-
-  // 输入提示符管理。
-  //
-  // readline 的 _refreshLine() 用 prevRows 追踪“提示符占了几行”,
-  // 下次刷新时 moveCursor(0, -prevRows) 上移光标。但模型输出直接写
-  // stdout,readline 不知道这些行,导致 prevRows 过期 → 光标错位。
-  //
-  // 修复:每次画提示符前,先写 \r\n 确保光标在新行第 0 列,
-  // 然后重置 prevRows=0 告诉 readline “当前行就是提示符行”。
-  // 这样 _refreshLine() 的 clearScreenDown() 只清当前空行,不影响模型输出。
-
-  const showPrompt = () => {
-    if (rl.closed) return
-    // 双保险:确保 stdin 处于 flowing 模式(ora 的 stdin-discarder 可能在
-    // stop() 时 pause 了 stdin,即使 discardStdin:false 已设,仍防止其他
-    // 库意外 pause)
-    if (process.stdin.isTTY && !process.stdin.readableFlowing) {
-      process.stdin.resume()
-    }
-    // bracketed paste 是终端层的开关,跑过的外部命令(全屏 TUI 之类)退出时可能把它关掉,
-    // 之后粘贴又会退回"每个 \n 一次提交"。它是幂等的,每次画提示符时重申一次。
-    enableBracketedPaste()
-    if (process.stdout.isTTY) {
-      // 确保光标在新行第 0 列(\r 回行首,\n 换行)
-      process.stdout.write('\r\n')
-      // 重置 readline 内部光标追踪
-      rl.prevRows = 0
-    }
-    rl.prompt()
-  }
-
-  // 安全版:防止 finally 块里的异常变成 unhandledRejection 杀掉会话
-  const safeShowPrompt = () => {
-    try {
-      showPrompt()
-    } catch (_) {
-      try { rl.prompt() } catch { /* 给予宽容 */ }
-    }
-  }
-
-  // 仅重画提示符,不写 \r\n(用于空输入、busy 等不需要换行的场景)
-  const safeRefreshPrompt = () => {
-    try {
-      if (rl.closed) return
-      enableBracketedPaste()
-      rl.prevRows = 0
-      rl.prompt()
-    } catch (_) {
-      try { rl.prompt() } catch { /* noop */ }
-    }
-  }
-
-  // 在提示符上方插一行通知(粘贴进度等)。
-  const notifyAbovePrompt = (text) => {
-    console.log(text)
-    if (!state.busy) safeShowPrompt()
-  }
-
-  // ── 斜杠命令即时提示(支持 ↑↓ 切换 / Tab、Enter 补全) ──
-  // 在提示符下方浮现匹配到的命令,随输入过滤;不移动光标(用 DECSC/DECRC 保存-恢复)。
-  // slashHintRows 记录当前提示占了几行,下次刷新/提交时据此擦除,避免残影。
-  // slashHintBaseLine:本次 hint 显示时"输入行的基线值" — ↑↓ 时 readline 已经把
-  //   rl.line 改成历史上一行,我们用基线回滚 + _refreshLine() 重绘,让用户视觉上看
-  //   不到"按 ↑ 输入行变了",只看到 hint 高亮跳了一格。
-  let slashHintRows = 0
-  let slashHintBaseLine = ''
-  let slashHintMatches = []
-  let slashHintIndex = 0
-  const eraseSlashHint = () => {
-    if (!process.stdout.isTTY || slashHintRows === 0) return
-    // 保存光标 → 下移到提示区逐行清空 → 恢复光标
-    let seq = '\x1b7'
-    for (let i = 0; i < slashHintRows; i++) seq += '\x1b[1B\x1b[2K'
-    seq += '\x1b8'
-    process.stdout.write(seq)
-    slashHintRows = 0
-  }
-  const drawSlashHint = () => {
-    const body = renderSlashHintBody(slashHintMatches, slashHintIndex, state.locale)
-    const hadVisibleHint = slashHintRows > 0
-    eraseSlashHint()
-    if (!body) return
-    const rows = body.split('\n')
-    if (!hadVisibleHint) {
-      // CSI B(光标下移)到达终端底部后只会停在最后一行,不会自动滚屏,
-      // 所以候选项会被裁掉。第一次显示提示时先用换行真实预留空间,
-      // 再上移相同行数回到输入行；无论中途是否发生滚屏,都能回到原内容所在行。
-      let reserve = ''
-      for (let i = 0; i < rows.length; i++) reserve += '\n\x1b[2K'
-      reserve += `\x1b[${rows.length}A`
-      process.stdout.write(reserve)
-    }
-    // 保存光标,逐行下移打印(每行先清行防止与旧内容叠字),最后恢复光标回输入行
-    let seq = '\x1b7'
-    for (const r of rows) seq += '\x1b[1B\x1b[2K\r' + r
-    seq += '\x1b8'
-    process.stdout.write(seq)
-    slashHintRows = rows.length
-  }
-  // 重置 hint 上下文:重新计算匹配 + 重画。输入字符变化时调用。
-  const refreshSlashHint = () => {
-    if (!process.stdout.isTTY || state.busy || state.inWizard) {
-      eraseSlashHint()
-      slashHintMatches = []
-      slashHintIndex = 0
-      slashHintBaseLine = ''
-      return
-    }
-    slashHintBaseLine = rl.line
-    slashHintMatches = filterSlashCommands(rl.line, state.locale)
-    // 匹配列表缩短时夹紧 index(增删交替时不会越界)
-    slashHintIndex = slashHintIndex < slashHintMatches.length ? slashHintIndex : 0
-    drawSlashHint()
-  }
-  // ↑↓ / Shift+Tab:readline 已经把 rl.line 改成历史行或没动 → 强制恢复基线并重绘输入行
-  const restoreInputLineToBaseline = () => {
-    if (rl.line === slashHintBaseLine) return
-    rl.line = slashHintBaseLine
-    rl.cursor = slashHintBaseLine.length
-    if (typeof rl._refreshLine === 'function') rl._refreshLine()
-  }
-  // Tab / Enter:把 rl.line 替换为补全文本 + 清掉 hint 状态。
-  // Enter 会先被 readline 提交,因此由下方 line handler 在处理命令前调用本函数。
-  const applyHintCompletion = (newLine) => {
-    rl.line = newLine
-    rl.cursor = newLine.length
-    eraseSlashHint()
-    slashHintMatches = []
-    slashHintIndex = 0
-    slashHintRows = 0
-    slashHintBaseLine = newLine
-    if (typeof rl._refreshLine === 'function') rl._refreshLine()
-  }
-
-  const neutralizeKeyForReadline = (key) => {
-    key.name = 'escape'
-    key.sequence = ''
-    key.ctrl = false
-    key.meta = false
-    key.shift = false
-  }
-
-  // readline 自己会优先处理 Enter/Ctrl+C。用 prependListener 在它之前完成补全
-  // 或取消当前操作,再把本次按键变成 readline 会忽略的 Esc。
-  const interceptInteractiveControlKeys = (_ch, key) => {
-    if (!key) return
-    if (key.ctrl && key.name === 'c') {
-      if (state.inWizard) {
-        neutralizeKeyForReadline(key)
-        rl.emit('operationCancel')
-        return
-      }
-      if (state.busy) {
-        neutralizeKeyForReadline(key)
-        state.cancelRequested = true
-        try { state.abortController?.abort() } catch (_) {}
-        void terminateCommand(state.currentChild)
-        return
-      }
-      if (slashHintRows > 0) {
-        neutralizeKeyForReadline(key)
-        applyHintCompletion('')
-        return
-      }
-      return
-    }
-    if (key.name !== 'return' && key.name !== 'enter') return
-    if (slashHintRows === 0 || slashHintMatches.length === 0 || state.busy || state.inWizard) return
-    const selected = slashHintMatches[slashHintIndex]
-    if (!selected) return
-    applyHintCompletion(selected.cmd + ' ')
-    neutralizeKeyForReadline(key)
-  }
-  rl.input.prependListener('keypress', interceptInteractiveControlKeys)
-
-  // Alt+V 粘贴剪贴板图片(node 把 ESC+v 解析为 meta+v;部分终端不发 Alt,可用 /image 兜底)
-  const pasteFromClipboard = async () => {
-    if (state.pasting) return
-    state.pasting = true
-    try {
-      notifyAbovePrompt(chalk.dim(t.imagePasting))
-      const img = await readClipboardImage()
-      if (!img) {
-        notifyAbovePrompt(chalk.yellow(t.imageEmpty))
-      } else {
-        state.pendingImages.push(img)
-        notifyAbovePrompt(chalk.green(t.imageAttached(state.pendingImages.length, formatBytes(img.bytes))))
-      }
-    } catch (err) {
-      // 不能让粘贴异常变成 unhandledRejection 杀掉会话
-      printError(`粘贴图片失败: ${err.message}`)
-    } finally {
-      state.pasting = false
-    }
-  }
-  rl.input.on('keypress', (ch, key) => {
-    if (!key) return
-
-    // ── 斜杠命令提示专用键盘:↑↓/Tab/Enter/Shift+Tab/Esc ──
-    // 仅在 hint 已显示且有匹配项时拦截,readline 的 default 行为(历史浏览等)
-    // 我们在 handler 里事后回滚 rl.line 来抵消
-    const action = parseKeyForSlashHint(key)
-    const hintActive = action === 'prev' || action === 'next' || action === 'complete' || action === 'cancel'
-      ? (slashHintRows > 0 && slashHintMatches.length > 0 && !state.busy && !state.inWizard)
-      : false
-
-    if (hintActive) {
-      if (action === 'prev' || action === 'next') {
-        const n = slashHintMatches.length
-        slashHintIndex = action === 'prev'
-          ? (slashHintIndex - 1 + n) % n
-          : (slashHintIndex + 1) % n
-        restoreInputLineToBaseline()
-        drawSlashHint()
-        return   // 不再走 default update + Alt+V 分支
-      }
-      if (action === 'complete') {
-        // Tab:把输入行替换为选中命令 + 末尾空格(便于继续打参数),
-        // 光标跳到末尾;之后不再触发 hint(已经精确选了某个命令)
-        const sel = slashHintMatches[slashHintIndex]
-        const newLine = sel.cmd + ' '
-        applyHintCompletion(newLine)
-        return
-      }
-      if (action === 'cancel') {
-        // Esc:只清 hint,不动输入行
-        eraseSlashHint()
-        slashHintMatches = []
-        slashHintIndex = 0
-        return
-      }
-    }
-
-    // Enter 走默认 readline 提交,line 事件里把选中项恢复到输入框;
-    // 其他普通按键:重算匹配 + 重画 hint
-    if (!(key.name === 'return' || key.name === 'enter')) {
-      refreshSlashHint()
-    }
-
-    // Alt+V 粘贴剪贴板图片(node 把 ESC+v 解析为 meta+v;部分终端不发 Alt,可用 /image 兜底)
-    const isAltV = (key.meta && key.name === 'v')
-      || (typeof key.sequence === 'string' && key.sequence.length === 2
-          && key.sequence.charCodeAt(0) === 27 && key.sequence[1] === 'v')
-    if (isAltV) pasteFromClipboard()
-  })
-
-  // 初始提示符:banner 后直接画(不需要 \r\n)
-  rl.prevRows = 0
-  rl.prompt()
-
-  rl.on('line', async (line) => {
-    // 提交时光标已落到提示符下一行(即提示区起点),直接清到屏幕末尾擦掉整块提示,
-    // 比 DECSC/DECRC 相对定位更稳(此时光标不在输入行)
-    if (process.stdout.isTTY && slashHintRows > 0) {
-      process.stdout.write('\x1b[0J')
-      slashHintRows = 0
-    }
-    // /addmodel 等交互式向导进行中时,用户的回答由向导自身的 rl.question 处理,
-    // 不应进入 REPL 的正常输入流程(否则会把向导的答案当成命令/消息发给模型)
-    if (state.inWizard) return
-    // 把输入行里的粘贴占位符还原成原文(带真实换行)—— 在判断是不是斜杠命令之前,
-    // 这样 `请写进 .env [粘贴 #1]` 这种混排也能正确展开
-    const pasted = pasteCapture.expand(line)
-    if (pasted.stale.length > 0) printWarn(t.pasteExpired(pasted.stale.length))
-    const input = pasted.text.trim()
-
-    if (input.startsWith('/')) {
-      const r = await handleSlashCommand(state, input, t)
-      if (r === 'exit') { rl.close(); return }
-      safeRefreshPrompt()
-      return
-    }
-
-    if (!input) { safeRefreshPrompt(); return }
-    if (state.busy) {
-      printDim(t.busy)
-      safeRefreshPrompt()
-      return
-    }
-
-    // 取出待发送图片(取出即清空队列,用户每条消息独立决定带不带图)
-    const images = state.pendingImages.splice(0)
-    if (images.length > 0) printDim(t.imageSending(images.length))
-
-    // 用了粘贴的这条消息:输入行里只有占位符,这里把真正发出去的内容亮一遍,
-    // 否则回看终端记录只剩一个 [粘贴 #N],等于把内容藏起来了。
-    // 只在 TTY 里回显 —— 管道/CI 下 readline 本来就不回显输入行(没有占位符这回事),
-    // 再补一块只会粘在提示符后面变成半行。
-    if (pasted.used > 0 && process.stdout.isTTY) process.stdout.write(renderPasteEcho(input, { locale: state.locale }))
-
-    state.busy = true
-    state.cancelRequested = false
-    state.abortController = new AbortController()
-    try {
-      await runAgentTurn(state, input, t, images)
-    } catch (err) {
-      if (state.cancelRequested) printDim(t.operationCancelled)
-      else printError(err.message)
-    } finally {
-      state.busy = false
-      state.cancelRequested = false
-      state.abortController = null
-      safeShowPrompt()
-    }
-  })
-
-  rl.on('close', () => {
-    if (state.busy) {
-      state.cancelRequested = true
-      state.abortController?.abort()
-      void terminateCommand(state.currentChild)
-    }
-    rl.input.removeListener('keypress', interceptInteractiveControlKeys)
-    pasteCapture.dispose()   // 卸载数据改写 + 把终端从 bracketed paste 模式里放出来
-    printWarn('\n' + t.bye)
-  })
-
-  // 全局安全网:任何未捕获的 Promise rejection / 异常都不应该静默杀掉
-  // 整个 REPL 会话。这里记录到 stderr 但不退出,让 readline 继续运行。
-  const rejectionHandler = (reason) => {
-    const msg = reason instanceof Error ? reason.message : String(reason)
-    printError(`内部错误(已捕获,会话继续): ${msg}`)
-    safeShowPrompt()
-  }
-  process.on('unhandledRejection', rejectionHandler)
-
-  // 等 readline 关闭后返回(进程随后自然退出)
-  await new Promise((resolve) => rl.on('close', resolve))
-
-  // 清理:移除本次会话注册的全局处理器
-  process.removeListener('unhandledRejection', rejectionHandler)
-}
-
-export default { runAiAgent }
+// Copyright 2026 xz333221
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+//
+// g ai — 终端交互式 AI 编码智能体。
+//
+// 入口: runAiAgent(argv)(由 gitCommit.js 在 argv[2]==='ai' 时动态 import)。
+//
+// 能力:
+//   - 默认使用 g ui 里配置的模型(~/.zen-gitsync/config.json 顶层 models,
+//     isDefault 优先,否则取第一个);--model=<序号|名称> 或会话内 /model 切换
+//   - OpenAI 兼容流式 function calling:模型可以调用 run_command /
+//     read_file / write_file / edit_file / list_files / search_text / ask_user
+//   - 图片:Alt+V 粘贴剪贴板图片、/image <路径> 附加本地图片,
+//     以多模态(image_url)消息发给模型(需模型支持视觉)
+//   - 权限:启动目录内全开放,其他目录可读写;唯一红线由 safety.js 拦截
+//     (格式化磁盘、删根目录、关机等系统级破坏操作)
+//
+// 视觉风格(termui.js):对标 Codex CLI / Claude Code / OpenCode —
+//   盒式输入框、ora 等待 spinner、✻ 思考(橙黄斜体)、⚙ 工具头(青) + │ 结果槽、
+//   正文 ➤ 子弹头(绿)+ 逐行轻量 markdown(**bold** / `code` / # 标题 / ``` 围栏 / - 列表)
+//
+// 使用方式:
+//   g ai                  交互模式(REPL)
+//   g ai "帮我修个 bug"    单发模式:执行一轮后退出
+//   g ai --model=2        指定配置里的第 2 个模型
+//
+// 实现注意:
+//   - 本模块属于 CLI 侧,不依赖 GUI 服务器代码(src/ui/server/**)
+//   - Ctrl+C 执行中只取消当前任务；空闲时交给主 CLI 清理并退出
+
+import readline from 'node:readline'
+import chalk from 'chalk'
+import config from '../../config.js'
+import { getCwd } from '../../utils/index.js'
+import { registerCleanup } from '../cleanup.js'
+import { terminateCommand, TOOL_DEFINITIONS } from './tools.js'
+import { runAgentTurn } from './turn.js'
+import { AgentExtensions } from './extensions.js'
+import { repairToolHistory, loadProjectInstructions, resolveRequestBudget } from './context.js'
+import {
+  printBanner, printHelpPanel,
+  filterSlashCommands, renderSlashHintBody, parseKeyForSlashHint,
+  printTurnSummary, renderPasteEcho,
+  printOk, printWarn, printError, printDim,
+} from './termui.js'
+import { readClipboardImage, checkImageFile, formatBytes } from './images.js'
+import { installPasteCapture, enableBracketedPaste, disableBracketedPaste } from './paste.js'
+import { runModelSetup, collectModelInput, buildModelConfig, selectFromList } from './modelSetup.js'
+import { genSessionId, autoTitle, writeSession, enforceRetention, listSessions } from './sessionStore.js'
+
+// truncateDisplay 已迁移到 termui.js;这里 re-export 保持既有测试/外部引用不断
+export { truncateDisplay } from './termui.js'
+
+// ──────────────────────────────────────────────
+// i18n 字符串表(CLI 其他部分以中文为主,这里按 locale 给双语)
+// ──────────────────────────────────────────────
+const STRINGS = {
+  zh: {
+    noModel: '未配置 AI 模型,请先在 GUI 通用设置中添加模型(运行 g ui 打开设置)',
+    unknownModel: (q) => `未找到匹配的模型: ${q},使用默认模型`,
+    waiting: '思考中…',
+    llmError: (msg) => `LLM 请求失败: ${msg}`,
+    toolIterLimit: (n) => `已达单轮最大工具调用次数(${n}),本轮结束。如需继续请再发一条消息。`,
+    toolRunning: (name) => `执行 ${name}…`,
+    busy: '任务执行中，请稍候，或按 Ctrl+C 停止当前任务。',
+    bye: '已退出 g ai',
+    cleared: '对话历史已清空',
+    newConversation: '已开启新对话',
+    operationCancelled: '已取消当前操作',
+    thinkOn: '思考过程显示: 开',
+    thinkOff: '思考过程显示: 关(模型仍会思考,只是不回显)',
+    modelSwitched: (label) => `已切换模型: ${label}`,
+    invalidModelIndex: '无效的模型序号',
+    cdOk: (p) => `工作目录已切换: ${p}`,
+    cdFail: (p) => `目录不存在: ${p}`,
+    currentModel: '当前模型',
+    availableModels: '可用模型',
+    oneShotNoModel: '未配置模型,无法启动',
+    oneShotDone: '单发模式完成 · 需要多轮连续对话请直接运行 g ai(不带参数)',
+    helpTitle: 'g ai 命令',
+    bannerTip: 'Alt+V 贴图 · /help 帮助 · Ctrl+C 停止当前任务 · /exit 退出',
+    prompt: '❯ ',
+    bannerModel: '模型',
+    bannerCwd: '目录',
+    thinkingLabel: '思考',
+    answerLabel: '回答',
+    thinkingHint: '… 已显示前 12 行；输入 /think full 开启后续完整思考',
+    emptyResponse: '模型没有返回回答，请重试或切换模型。',
+    chars: '字符',
+    extLoadingMcp: (n) => `正在连接 ${n} 个 MCP 服务…`,
+    extLoaded: (skills, tools) => `扩展已就绪: ${skills} 个 Skill · ${tools} 个 MCP 工具(/skills 查看)`,
+    extNone: '',
+    extSkillTitle: '已安装的 Skill',
+    extSkillEmpty: '没有装 Skill。可以在 g ui → 智能体 → Skill 广场 里挑一个装到当前项目或 g ai。',
+    extMcpTitle: '已接入的 MCP 服务',
+    extMcpEmpty: '没有接入 MCP 服务。可以在 g ui → 智能体 → MCP 广场 里挑一个装到当前项目或 g ai。',
+    extMcpFailed: '连接失败',
+    imagePasting: '正在读取剪贴板图片…',
+    imageAttached: (n, size) => `📎 图片 #${n} 已附加(${size}),将随下一条消息发送`,
+    imageEmpty: '剪贴板中没有图片;可先截图再按 Alt+V,或用 /image <路径> 附加本地图片',
+    imageBadPath: (p) => `图片不存在或格式不支持: ${p}(支持 png/jpg/jpeg/gif/webp/bmp)`,
+    imageListTitle: '待发送图片',
+    imageListEmpty: '当前没有待发送的图片(Alt+V 粘贴或 /image <路径> 附加)',
+    imageCleared: '已清除待发送的图片',
+    imageSending: (n) => `📎 附带 ${n} 张图片`,
+    pasteToken: (n, lines) => `[粘贴 #${n} · ${lines} 行]`,
+    pasteExpired: (n) => `有 ${n} 段粘贴内容已从内存回收，无法还原；如仍需要请重新粘贴`,
+    addModelTitle: '添加模型配置',
+    addModelSaved: (name) => `✓ 模型 "${name}" 已添加并切换为当前模型`,
+    addModelCancelled: '已取消添加模型',
+    addModelSaveError: (msg) => `保存配置失败: ${msg}`,
+    resumeTitle: '选择要恢复的对话',
+    resumeEmpty: '没有可恢复的历史对话',
+    resumeCancelled: '已取消恢复对话',
+    resumeDone: (title, count) => `已恢复对话: ${title} (${count} 条消息)`,
+  },
+  en: {
+    noModel: 'No AI model configured. Add one in GUI settings first (run `g ui`).',
+    unknownModel: (q) => `No model matching "${q}", falling back to default`,
+    waiting: 'Thinking…',
+    llmError: (msg) => `LLM request failed: ${msg}`,
+    toolIterLimit: (n) => `Hit max tool iterations (${n}) for this turn. Send another message to continue.`,
+    toolRunning: (name) => `Running ${name}…`,
+    busy: 'Agent is working. Wait or press Ctrl+C to stop the task.',
+    bye: 'Bye',
+    cleared: 'Conversation cleared',
+    newConversation: 'Started a new conversation',
+    operationCancelled: 'Current operation cancelled',
+    thinkOn: 'Thinking display: on',
+    thinkOff: 'Thinking display: off (model still thinks, just hidden)',
+    modelSwitched: (label) => `Switched model: ${label}`,
+    invalidModelIndex: 'Invalid model index',
+    cdOk: (p) => `Working directory changed: ${p}`,
+    cdFail: (p) => `Directory not found: ${p}`,
+    currentModel: 'Current model',
+    availableModels: 'Available models',
+    oneShotNoModel: 'No model configured, aborting',
+    oneShotDone: 'One-shot done · for multi-turn chat run `g ai` with no arguments',
+    helpTitle: 'g ai commands',
+    bannerTip: 'Alt+V image · /help commands · Ctrl+C stop task · /exit quit',
+    prompt: '❯ ',
+    bannerModel: 'Model',
+    bannerCwd: 'CWD',
+    thinkingLabel: 'Thinking',
+    answerLabel: 'Answer',
+    thinkingHint: '… first 12 lines shown; enter /think full for future reasoning',
+    emptyResponse: 'The model returned no answer. Retry or switch models.',
+    chars: 'chars',
+    extLoadingMcp: (n) => `Connecting to ${n} MCP server(s)…`,
+    extLoaded: (skills, tools) => `Extensions ready: ${skills} skill(s) · ${tools} MCP tool(s) (/skills for details)`,
+    extNone: '',
+    extSkillTitle: 'Installed skills',
+    extSkillEmpty: 'No skills installed. Browse g ui → Agent → Skill marketplace to install one into this project or g ai.',
+    extMcpTitle: 'Connected MCP servers',
+    extMcpEmpty: 'No MCP servers connected. Browse g ui → Agent → MCP marketplace to install one.',
+    extMcpFailed: 'failed',
+    imagePasting: 'Reading clipboard image…',
+    imageAttached: (n, size) => `📎 Image #${n} attached (${size}); sent with your next message`,
+    imageEmpty: 'No image in clipboard; take a screenshot first, or use /image <path>',
+    imageBadPath: (p) => `Not a supported image: ${p} (png/jpg/jpeg/gif/webp/bmp)`,
+    imageListTitle: 'Pending images',
+    imageListEmpty: 'No pending images (Alt+V to paste, or /image <path>)',
+    imageCleared: 'Pending images cleared',
+    imageSending: (n) => `📎 ${n} image(s) attached`,
+    pasteToken: (n, lines) => `[paste #${n} · ${lines} lines]`,
+    pasteExpired: (n) => `${n} pasted block(s) were dropped from memory and could not be restored; paste again if still needed`,
+    addModelTitle: 'Add Model Configuration',
+    addModelSaved: (name) => `✓ Model "${name}" added and switched to`,
+    addModelCancelled: 'Add model cancelled',
+    addModelSaveError: (msg) => `Failed to save: ${msg}`,
+    resumeTitle: 'Choose a conversation to resume',
+    resumeEmpty: 'No previous conversations to resume',
+    resumeCancelled: 'Resume cancelled',
+    resumeDone: (title, count) => `Resumed: ${title} (${count} messages)`,
+  },
+}
+
+function makeStrings(locale) {
+  return String(locale || '').startsWith('en') ? STRINGS.en : STRINGS.zh
+}
+
+// 模型的展示名:优先 model 字段,其次 name,最后 baseURL
+function modelLabel(m) {
+  return m?.model || m?.name || m?.baseURL || '(unknown)'
+}
+
+// ──────────────────────────────────────────────
+// 系统 prompt — 明确告知环境、权限边界与工作方式
+// ──────────────────────────────────────────────
+// 顺序固定:基础人设 → 项目说明(AGENTS.md/CLAUDE.md)→ 扩展能力(skill 清单 + MCP 工具)。
+// 把扩展放最后,是因为它是"增量能力",不该插在权限与环境说明中间打断阅读。
+// state 里存着已加载的 AgentExtensions;没加载(如单测直接调)时 extra 为空,行为与改动前一致。
+function extensionSuffix(state) {
+  try {
+    return state?.extensions?.promptSuffix?.(state.locale) || ''
+  } catch {
+    return ''
+  }
+}
+
+async function buildProjectPrompt(options) {
+  return buildSystemPrompt(options) + await loadProjectInstructions(options.cwd) + (options.extra || '')
+}
+
+function buildSystemPrompt({ cwd, locale, shellDesc }) {
+  const zh = !String(locale || '').startsWith('en')
+  const now = new Date().toLocaleString()
+  const isWin = process.platform === 'win32'
+  // 工具数量动态取自 tools.js —— 之前这里写死过"6 个",加 ask_user 时没人想起来改,
+  // 就变成了对模型撒谎。数字跟着定义走,以后再加工具不会漏。
+  const builtinToolCount = TOOL_DEFINITIONS.length
+  if (zh) {
+    return `你是 "g ai" —— zen-gitsync CLI 内置的终端编码智能体,通过工具在用户真实电脑上完成编码任务。
+
+# 运行环境
+- 操作系统: ${process.platform}
+- Shell: ${shellDesc}
+- 当前工作目录: ${cwd}(用户在此启动 g ai,也是所有相对路径的基准)
+- 当前时间: ${now}
+
+# 平台兼容性(重要!)
+- 必须使用与当前 Shell 兼容的命令,禁止盲套 Unix 写法
+${isWin ? `- 当前是 Windows cmd.exe,以下 Unix 命令**不存在**,用了必定报"不是内部或外部命令":
+  tail / head / cat / grep / ls / sed / awk / wc / cut / uniq / xargs / which / touch
+  平台守卫会在执行前拦截这些命令,但请主动避免,不要浪费一轮调用
+- 跨平台替代方案:
+  · 列目录 → list_files 工具 或 cmd 的 dir
+  · 搜内容 → search_text 工具 或 cmd 的 findstr
+  · 看文件 → read_file 工具 或 cmd 的 type
+  · 看图片 → read_image 工具(截图/设计稿/示意图)。read_file 读图片只会得到乱码,别浪费那一轮
+  · 看输出末尾 → PowerShell "命令 | Select-Object -Last N"
+  · 文本处理 → node -e "..." 或 PowerShell
+  · 查命令路径 → cmd 的 where(不是 which)
+- 必须跑 shell 时优先跨平台写法(如 node -e "..."),别用 Unix 专属命令` : `- 当前是 POSIX 环境,Unix 命令可用`}
+
+# 远程仓库(GitHub / Gitee)
+- 问"这个项目关联哪个远端" → run_command 跑 \`git remote -v\`:本地信息,不联网、不依赖任何 CLI,任何环境都能答
+- 问"我账号下有哪些仓库"或某仓库的 PR / Issue → 用官方 CLI。凭据由 CLI 自己保管:
+  · GitHub → \`gh\`。列仓库 \`gh repo list --limit 50 --json name,visibility,updatedAt,primaryLanguage\`;
+    看详情 \`gh repo view <owner/repo>\`;PR \`gh pr list\`;登录态 \`gh auth status\`
+  · Gitee → \`gitee\`。列仓库 \`gitee repo list\`;登录态 \`gitee auth status\`
+    (未登录时退出码**仍是 0**,必须看 \`--json\` 里的 status 字段,只看退出码会误判成已登录)
+  · 两者默认只覆盖**当前登录账号**;组织仓库、私有仓库可能需要更大的 token scope,拉不到就如实说明,
+    不要用其他途径绕过
+  · 绝不向用户索要 token —— 也不要让用户把 token 粘进对话
+- CLI 可能没装、或没在当前进程的 PATH 里(装完没重启服务就是这个表现):
+  报"不是内部或外部命令" / command not found 时,**不要换写法反复重试**,更不要凭印象编仓库名或目录名。
+  直接告诉用户未检测到该 CLI,可在 ZenGitSync 的「远程仓库」页一键安装/登录;
+  若只是想回答本项目的问题,退回 \`git remote -v\`
+- 克隆仓库 / 添加远端时**优先用 SSH**(用户偏好,本机已配好密钥,实测走 https 会弹凭据窗口把任务打断):
+  · GitHub 用 \`git@github.com:owner/repo.git\`,Gitee 用 \`git@gitee.com:owner/repo.git\`
+  · 用户给的是 \`https://...\` 或 \`gh repo clone owner/repo\` 时,先换算成 SSH 地址再执行 ——
+    走 https 会弹 Git Credential Manager 让用户输账号密码,任务就停在半路等输入
+  · 已有仓库换协议:\`git remote set-url origin <ssh 地址>\`(先 \`git remote -v\` 看当前是什么)
+  · 只有 SSH 真的不可用(报 \`Permission denied (publickey)\` / \`Host key verification failed\`)才退回 https,
+    并用一句话说明这次走的是 https、配好密钥后可改回;两边都失败就停下来问用户,不要反复重试
+
+# 权限(用户已明确授权,无需反复征求同意)
+- 工作目录内:读写文件、执行命令等所有操作直接执行
+- 其他目录:同样可以读取和修改
+- 唯一红线:不得破坏系统(格式化磁盘、删除根目录/系统目录、关机重启、写块设备等)。
+  命令守卫只能识别部分危险写法，不是沙箱；不得通过脚本或嵌套 shell 绕过限制。
+
+# 工作方式
+- 先动手、后提问:能用工具查清的不要问用户(list_files / read_file / search_text / run_command)
+- 修改代码后主动验证:跑测试、构建或至少语法检查(用 run_command)
+- 编辑文件优先 edit_file 精确替换;先 read_file 看原文,old_string 必须与文件内容完全一致(含缩进换行)
+- 大文件用 offset/limit 分段读取,不要一次读爆上下文
+- git 操作用 run_command 执行;提交代码可以用 git 命令,也可以用本 CLI 的 g -y(默认信息提交并推送)或 g --ai(AI 生成提交信息)
+- run_command 默认就在工作目录执行,不要再加 cd / cd /d 前缀;默认超时 120 秒,长任务加大 timeout_seconds(最大 600)
+- 命令在 ${shellDesc} 下执行,注意语法兼容
+- 先读取当前项目的规范、依赖与测试配置，再选择对应的验证命令，不要假定所有项目都使用 npm
+- 用户要求概述或简短回答时，只读取必要的结构与入口文件，控制探索范围
+
+# 任务计划(多步任务必用)
+- 任务需要多步时(改代码、排查、调研、多文件改动),**动手之前**先调一次 update_plan,把要做的事
+  拆成 3-8 个可核对的步骤,让用户在你改任何东西之前就知道范围和验收标准
+- 单步小事(读个文件、答一个问题)不要列计划,那是噪音
+- 每次都传**完整**的当前计划(不是增量),顺序即执行顺序;同一时刻最多一个 in_progress
+- 完成一步就更新一次状态,别攒到最后一次性刷;计划与实际不符时直接改计划,不要硬着头皮往下走
+- 全部做完后把 steps 传空表示收尾
+- update_plan 只是给人看的进度板,不要为了"更新计划"去改文件或跑命令
+
+# 与用户交互
+- 需要向用户确认、提问或汇报重要决策时,直接用普通文本输出 —— 用户能实时看到你的文本;
+  需要暂停当前任务并等待用户决定或补充信息时,调用 ask_user,不要猜测或只在普通文本里提问
+  不要调用不存在的工具,内置工具就是上面列出的 ${builtinToolCount} 个(若用户装了 MCP,你的工具表里还会多出 mcp__ 开头的工具)
+- 图片有两条来路:
+  · 用户通过 Alt+V 或 /image 附加 → 图片以 image_url 部件出现在 user 消息里,直接就能看到
+  · 用户只给了**路径**(或你自己在翻文件时遇到图)→ 用 read_image 工具读它,别用 read_file
+    (read_file 按 UTF-8 解码,图片只会是乱码),也别根据文件名猜内容
+  如果当前模型不支持视觉(带图请求报错),提醒用户换用支持视觉的模型
+- 发现高风险或状态不一致的情况(例如版本号 / git tag / CHANGELOG 对不上、发布前环境异常、
+  仓库状态与预期不符)时:先用文本说明发现和影响,停下来等用户指示,不要擅自继续破坏性操作
+
+# 输出
+- 你的文本输出直接显示在用户终端,用简体中文交流
+- 完成任务后用一两句话汇报结果,不要复述过程细节`
+  }
+  return `You are "g ai" — the terminal coding agent built into the zen-gitsync CLI. You use tools to perform coding tasks on the user's real machine.
+
+# Environment
+- OS: ${process.platform}
+- Shell: ${shellDesc}
+- Working directory: ${cwd} (where the user launched g ai; base for all relative paths)
+- Current time: ${now}
+
+# Platform compatibility (important!)
+- You MUST use commands compatible with the current shell. Do NOT blindly copy Unix patterns.
+${isWin ? `- This is Windows cmd.exe. The following Unix commands do NOT exist here and will fail with "not recognized":
+  tail / head / cat / grep / ls / sed / awk / wc / cut / uniq / xargs / which / touch
+  A platform guard will block these before execution, but avoid them proactively — don't waste a turn.
+- Cross-platform alternatives:
+  · List dirs → list_files tool, or cmd's dir
+  · Search content → search_text tool, or cmd's findstr
+  · Read files → read_file tool, or cmd's type
+  · Look at an image → read_image tool (screenshots, mockups, diagrams). read_file only decodes text and returns garbage for images
+  · Tail output → PowerShell "command | Select-Object -Last N"
+  · Text processing → node -e "..." or PowerShell
+  · Find executable → cmd's where (not which)
+- For shell one-offs, prefer cross-platform forms like node -e "..."` : `- POSIX environment: Unix commands are available`}
+
+# Remote repositories (GitHub / Gitee)
+- "Which remote does this project point at?" → run_command \`git remote -v\`: local info, no network, no CLI needed
+- "Which repos do I have?" or a repo's PRs / issues → use the official CLI. The CLI owns the credentials:
+  · GitHub → \`gh\`. List repos \`gh repo list --limit 50 --json name,visibility,updatedAt,primaryLanguage\`;
+    details \`gh repo view <owner/repo>\`; PRs \`gh pr list\`; auth state \`gh auth status\`
+  · Gitee → \`gitee\`. List repos \`gitee repo list\`; auth state \`gitee auth status\`
+    (the exit code is still 0 when logged out — read the status field from \`--json\`, or you will
+    wrongly report "logged in")
+  · Both default to the **currently authenticated account only**; org and private repos may need
+    broader token scopes. If you cannot see them, say so plainly — do not route around it
+  · Never ask the user for a token, and never have them paste one into the conversation
+- The CLI may not be installed, or missing from this process's PATH (that is what "installed but the
+  server cannot find it" looks like):
+  on "not recognized" / command not found, do NOT retry with a different spelling and do NOT invent
+  repo or directory names. Tell the user the CLI was not found and point them to ZenGitSync's
+  "Remote repositories" page (one-click install / login); if you only need to answer for this project,
+  fall back to \`git remote -v\`
+- Cloning a repo or adding a remote: **prefer SSH** (the user's preference — keys are already set up,
+  and https has been observed to pop a credential window that stalls the task):
+  · GitHub → \`git@github.com:owner/repo.git\`; Gitee → \`git@gitee.com:owner/repo.git\`
+  · If the user hands you an \`https://...\` URL (or \`gh repo clone owner/repo\`), convert it to the SSH
+    form first — https pops up Git Credential Manager asking for a username/password and blocks the task
+  · Switching an existing repo: \`git remote set-url origin <ssh url>\` (check \`git remote -v\` first)
+  · Fall back to https only when SSH genuinely fails (\`Permission denied (publickey)\` /
+    \`Host key verification failed\`), and say in one line that this one used https and can go back to SSH
+    once the key is set up; if both fail, stop and ask the user — do not keep retrying
+
+# Permissions (explicitly granted by the user — do not keep asking)
+- Inside the working directory: read/write files and run commands directly
+- Other directories: may also be read and modified
+- Never destroy the system (format disks, delete root/system dirs, shutdown/reboot, write block devices). The command guard is not a sandbox; never bypass it with scripts or nested shells.
+
+# How to work
+- Act first, ask later: use tools (list_files / read_file / search_text / run_command) instead of asking the user
+- When a decision or missing detail must come from the user, call ask_user. Do not guess or merely describe a question in plain text.
+- After editing code, verify: run tests, builds, or at least a syntax check via run_command
+- Prefer edit_file for precise replacements; read_file first, old_string must match the file exactly
+- Read large files in segments (offset/limit)
+- Git operations go through run_command; to commit, use git commands or this CLI's g -y / g --ai
+- run_command already executes in the working directory — do NOT prefix with cd; default timeout 120s, raise timeout_seconds (max 600) for long tasks
+- Commands run under ${shellDesc}; keep syntax compatible
+- Read this project's instructions, dependencies and test configuration before choosing verification commands; do not assume npm
+- For a brief overview, inspect only the necessary structure and entry points; keep exploration proportional to the request
+
+# Task plan (required for multi-step work)
+- When a task takes several steps (code changes, debugging, research, multi-file edits), call update_plan
+  **before touching anything**: break it into 3-8 verifiable steps so the user knows the scope and the
+  acceptance criteria before you modify a single file
+- Do not plan single trivial actions (read one file, answer one question) — that is noise
+- Always send the **complete** current plan (not a delta), in execution order; at most one in_progress
+- Update the status right after finishing a step instead of batching everything at the end; when the
+  plan no longer matches reality, rewrite the plan instead of pushing ahead
+- Send an empty steps list once everything is done
+- update_plan is a progress board for humans: never edit files or run commands just to "update the plan"
+
+# Talking to the user
+- When you need to confirm something, ask a question, or report an important decision, just write plain text — the user sees your output in real time. Use ask_user when the task must pause for an answer. Never call tools that do not exist; the built-in set is the ${builtinToolCount} tools listed above (if the user installed MCP servers, extra tools prefixed with mcp__ will also appear in your tool list)
+- Images reach you two ways:
+  · The user attached one via Alt+V or /image → it arrives as an image_url part in the user message; you can see it directly
+  · The user only gave you a **path** (or you run into an image while exploring) → read it with the read_image tool, not read_file
+    (read_file decodes UTF-8 and returns garbage for images), and never guess an image's content from its filename
+  If the current model rejects images (no vision support), tell the user to switch to a vision-capable model
+- When you spot high-risk or inconsistent state (version number / git tag / CHANGELOG mismatch, abnormal release environment, unexpected repo state), explain the finding and its impact in text, then STOP and wait for the user's decision instead of proceeding with destructive operations
+
+# Output
+- Your text output goes straight to the user's terminal; reply in English
+- Report results in one or two sentences; do not narrate the process`
+}
+
+// ──────────────────────────────────────────────
+// 消息准备(消毒 / 旧图片降级)统一实现在 context.js —— CLI 与 Web 智能体面板
+// 共用同一份,避免两侧各写一遍后行为分叉。这里 re-export 保持既有引用与测试不断。
+// ──────────────────────────────────────────────
+export { sanitizeMessages, stripStaleImages } from './context.js'
+
+// ──────────────────────────────────────────────
+// 单轮 agent 循环:用户一句话 → 流式输出 → 工具调用 → 再调用模型 … 直到模型给出最终文本
+// images: [{path, bytes}] 待发送图片(可为空数组)
+// ──────────────────────────────────────────────
+function printSlashHelp(t, locale) {
+  const zh = !String(locale || '').startsWith('en')
+  const lines = zh ? [
+    '  /help             显示本帮助',
+    '  /model            列出可用模型;/model <序号> 切换模型',
+    '  /addmodel         添加新的模型配置(交互式向导)',
+    '  /cd <路径>        切换智能体工作目录',
+    '  /image [路径]     查看待发送图片;/image <路径> 附加本地图片;/image clear 清除',
+    '  /think full       完整显示后续思考（默认，模型返回的内容）',
+    '  /think compact    预览前 12 行思考',
+    '  /think off        隐藏思考；/think 切换完整与隐藏',
+    '  /tools [模式]     compact 精简 / full 完整工具输出',
+    '  /skills           查看已安装的 Skill 与已接入的 MCP 服务',
+    '  /stats            查看上一轮与会话累计用量、耗时',
+    '  /new              开启新对话',
+    '  /resume           选择并恢复之前的对话',
+    '  /clear            清空对话历史',
+    '  /exit, /quit      退出',
+    '',
+    '  Alt+V             粘贴剪贴板图片,随下一条消息发送(需视觉模型)',
+    '  多行粘贴          直接粘贴多行文本:整段作为一条消息发送,输入行里显示 [粘贴 #N] 占位符',
+    '',
+    '权限说明: 工作目录内全部操作直接执行;其他目录可读写;',
+    '命令守卫识别部分系统级危险操作，不是隔离沙箱。',
+  ] : [
+    '  /help             Show this help',
+    '  /model            List models; /model <n> to switch',
+    '  /addmodel         Add a new model (interactive wizard)',
+    '  /cd <path>        Change agent working directory',
+    '  /image [path]     List pending images; attach a file; /image clear to reset',
+    '  /think full       Show future thinking returned by the model in full (default)',
+    '  /think compact    Preview the first 12 thinking lines',
+    '  /think off        Hide thinking; /think toggles full / hidden',
+    '  /tools [mode]     compact / full tool output',
+    '  /skills           List installed skills and connected MCP servers',
+    '  /stats            Last turn timing and session token usage',
+    '  /new              Start a new conversation',
+    '  /resume           Choose and resume a previous conversation',
+    '  /clear            Clear conversation history',
+    '  /exit, /quit      Quit',
+    '',
+    '  Alt+V             Paste clipboard image (needs a vision-capable model)',
+    '  Multi-line paste  Paste a block as one message; the input line shows a [paste #N] placeholder',
+    '',
+    'Permissions: full access in the working directory; other dirs readable/writable;',
+    'the command guard detects some dangerous operations; it is not a sandbox.',
+  ]
+  printHelpPanel(t.helpTitle, lines)
+}
+
+export async function handleSlashCommand(state, input, t) {
+  const [cmd, ...rest] = input.split(/\s+/)
+  const arg = rest.join(' ').trim()
+  const zh = !String(state.locale).startsWith('en')
+  if (state.busy && !['/help', '/stats', '/think', '/tools', '/skills', '/mcp', '/exit', '/quit'].includes(cmd)) {
+    printDim(t.busy)
+    return 'ok'
+  }
+
+  if (cmd === '/exit' || cmd === '/quit') return 'exit'
+  if (cmd === '/help') { printSlashHelp(t, state.locale); return 'ok' }
+  if (cmd === '/clear' || cmd === '/new') {
+    await state.persistSession?.()
+    state.messages = [state.messages[0]]
+    // 开新会话 ID,旧会话保留在磁盘上
+    state.sessionId = genSessionId()
+    state.sessionCreatedAt = new Date().toISOString()
+    state.lastTurnStats = null
+    state.sessionStats = null
+    printOk(cmd === '/new' ? t.newConversation : t.cleared)
+    return 'ok'
+  }
+  if (cmd === '/resume') {
+    if (state.busy) {
+      printDim(t.busy)
+      return 'ok'
+    }
+    const sessions = (await listSessions())
+      .filter(session => session.sessionId !== state.sessionId && session.messages.length > 1)
+    if (sessions.length === 0) {
+      printDim(t.resumeEmpty)
+      return 'ok'
+    }
+
+    const zh = !String(state.locale || '').startsWith('en')
+    const cleanLabel = (value, fallback) => String(value || fallback).replace(/[\r\n\x00-\x1f\x7f]/g, ' ').trim()
+    const items = sessions.map((session) => {
+      const title = cleanLabel(session.title, zh ? '(未命名对话)' : '(untitled)')
+      const updated = new Date(session.updatedAt || session.createdAt || 0)
+      const date = Number.isNaN(updated.getTime()) ? '' : updated.toLocaleString(state.locale)
+      const count = session.messages.filter(message => message?.role !== 'system').length
+      return {
+        label: `${title}  ${chalk.dim(`${date} · ${count} ${zh ? '条消息' : 'messages'}`)}`,
+        value: session,
+      }
+    })
+    const selectStrings = {
+      navHint: zh ? '↑↓ 切换,Enter 确认' : '↑↓ navigate, Enter confirm',
+      selectPrompt: (min, max) => zh ? `请选择(${min}-${max}):` : `Choose (${min}-${max}):`,
+      invalidChoice: (min, max) => zh ? `请输入 ${min}-${max} 之间的序号` : `Enter a number from ${min} to ${max}`,
+    }
+    const asker = {
+      ask(prompt) {
+        return new Promise(resolve => state.rl.question(`${prompt} `, answer => resolve(String(answer || '').trim())))
+      },
+    }
+
+    state.inWizard = true
+    try {
+      const selected = await selectFromList({
+        rl: state.rl,
+        asker,
+        title: t.resumeTitle,
+        items,
+        t: selectStrings,
+      })
+      if (!selected?.value) return 'ok'
+
+      const session = selected.value
+      const { promises: fsp } = await import('node:fs')
+      if (session.cwd) {
+        try {
+          const stat = await fsp.stat(session.cwd)
+          if (stat.isDirectory()) state.ctx.cwd = session.cwd
+        } catch { /* 已删除的旧目录不阻断恢复,沿用当前目录 */ }
+      }
+      state.sessionId = session.sessionId
+      state.sessionCreatedAt = session.createdAt || new Date().toISOString()
+      state.messages = repairToolHistory(session.messages)
+      state.lastTurnStats = session.lastTurnStats || null
+      state.sessionStats = session.sessionStats || null
+      const systemMessage = {
+        role: 'system',
+        content: await buildProjectPrompt({ cwd: state.ctx.cwd, locale: state.locale, shellDesc: state.shellDesc, extra: extensionSuffix(state) }),
+      }
+      if (state.messages[0]?.role === 'system') state.messages[0] = systemMessage
+      else state.messages.unshift(systemMessage)
+      state.pendingImages = []
+      const count = state.messages.filter(message => message?.role !== 'system').length
+      printOk(t.resumeDone(cleanLabel(session.title, session.sessionId), count))
+    } catch (err) {
+      if (err?.cancelled) printDim(t.resumeCancelled)
+      else printError(err.message)
+    } finally {
+      state.inWizard = false
+    }
+    return 'ok'
+  }
+  if (cmd === '/think') {
+    const next = arg || (state.showThinking ? 'off' : 'full')
+    if (!['off', 'compact', 'full'].includes(next)) {
+      printWarn('/think compact | full | off')
+      return 'ok'
+    }
+    state.showThinking = next !== 'off'
+    state.thinkingMode = next
+    printOk(zh ? `后续请求的思考显示：${({ off: '隐藏', compact: '预览前 12 行', full: '完整' })[next]}` : `Thinking for future requests: ${next}`)
+    return 'ok'
+  }
+  if (cmd === '/tools') {
+    if (arg && !['full', 'compact'].includes(arg)) { printWarn('/tools compact | full'); return 'ok' }
+    state.fullTools = arg ? arg === 'full' : !state.fullTools
+    printOk(zh ? `后续工具结果：${state.fullTools ? '完整' : '精简'}` : `Future tool output: ${state.fullTools ? 'full' : 'compact'}`)
+    return 'ok'
+  }
+  if (cmd === '/skills' || cmd === '/mcp') {
+    const info = state.extensions?.summary(state.locale) || { skills: [], servers: [] }
+    console.log(chalk.bold(`\n  ${t.extSkillTitle}`))
+    if (info.skills.length === 0) printDim(t.extSkillEmpty)
+    for (const skill of info.skills) {
+      const scope = skill.scope === 'project' ? (zh ? '项目' : 'project') : (zh ? '全局' : 'global')
+      console.log(`  · ${chalk.cyan(skill.name)} ${chalk.dim(`[${scope}]`)}${skill.description ? chalk.dim(` — ${skill.description}`) : ''}`)
+    }
+    console.log(chalk.bold(`\n  ${t.extMcpTitle}`))
+    if (info.servers.length === 0) printDim(t.extMcpEmpty)
+    for (const server of info.servers) {
+      if (server.error) {
+        console.log(`  · ${chalk.red(server.id)} ${chalk.dim(`— ${t.extMcpFailed}: ${server.error}`)}`)
+      } else {
+        console.log(`  · ${chalk.cyan(server.id)} ${chalk.dim(`— ${server.tools} ${zh ? '个工具' : 'tools'} (${server.command})`)}`)
+      }
+    }
+    console.log('')
+    return 'ok'
+  }
+  if (cmd === '/stats') {
+    if (state.lastTurnStats) printTurnSummary(state.lastTurnStats, { locale: state.locale, session: state.sessionStats })
+    else printDim(zh ? '完成一轮对话后显示用量与耗时。' : 'Usage and timing appear after a turn completes.')
+    return 'ok'
+  }
+  if (cmd === '/image') {
+    if (arg === 'clear') {
+      state.pendingImages = []
+      printOk(t.imageCleared)
+      return 'ok'
+    }
+    if (!arg) {
+      if (state.pendingImages.length === 0) {
+        printDim(t.imageListEmpty)
+      } else {
+        console.log(chalk.bold(t.imageListTitle) + ':')
+        state.pendingImages.forEach((img, i) => {
+          console.log(`  ${i + 1}. ${img.path} ${chalk.dim('(' + formatBytes(img.bytes) + ')')}`)
+        })
+      }
+      return 'ok'
+    }
+    const { resolve } = await import('node:path')
+    const target = resolve(state.ctx.cwd, arg)
+    const img = await checkImageFile(target)
+    if (!img) {
+      printError(t.imageBadPath(target))
+      return 'ok'
+    }
+    state.pendingImages.push(img)
+    printOk(t.imageAttached(state.pendingImages.length, formatBytes(img.bytes)))
+    return 'ok'
+  }
+  if (cmd === '/model') {
+    if (!arg) {
+      console.log(chalk.bold(t.availableModels) + ':')
+      state.models.forEach((m, i) => {
+        const current = m === state.model ? chalk.green(' ← ' + t.currentModel) : ''
+        const keyHint = m.apiKey ? '' : chalk.yellow(' (无 apiKey)')
+        console.log(`  ${i + 1}. ${modelLabel(m)} ${chalk.dim(m.baseURL || '')}${keyHint}${current}`)
+      })
+      console.log(chalk.dim('/model <序号> 切换'))
+      return 'ok'
+    }
+    const idx = Number.parseInt(arg, 10)
+    if (!Number.isFinite(idx) || idx < 1 || idx > state.models.length) {
+      printError(t.invalidModelIndex)
+      return 'ok'
+    }
+    state.model = state.models[idx - 1]
+    printOk(t.modelSwitched(modelLabel(state.model)))
+    return 'ok'
+  }
+  if (cmd === '/addmodel') {
+    // 智能体执行中不允许启动向导:向导的 rl.question 会与流式输出交错,体验混乱
+    if (state.busy) {
+      printDim(t.busy)
+      return 'ok'
+    }
+    state.inWizard = true
+    try {
+      console.log(chalk.cyan.bold('\n' + t.addModelTitle))
+      // 复用 modelSetup 的交互式收集逻辑(服务商 → 接口 → 模型名 → Key → 显示名 → 测试)
+      const collected = await collectModelInput({
+        locale: state.locale,
+        rl: state.rl,
+        cancelMessage: t.addModelCancelled,
+      })
+      if (!collected) return 'ok'
+      const { baseURL, model, apiKey, displayName } = collected
+      // 构建模型配置(isDefault=false:/addmodel 只是追加,不改默认模型)
+      const newModel = buildModelConfig({ baseURL, model, apiKey, name: displayName, isDefault: false })
+      // 持久化到 ~/.zen-gitsync/config.json 顶层 models 数组
+      const cfg = await config.loadConfig()
+      const models = Array.isArray(cfg.models) ? cfg.models : []
+      models.push(newModel)
+      cfg.models = models
+      try {
+        await config.saveConfig(cfg)
+      } catch (err) {
+        printError(t.addModelSaveError(err.message))
+        return 'ok'
+      }
+      // 同步 REPL 状态:更新模型列表 + 自动切换到新模型(用户刚添加,大概率想立刻用)
+      state.models = models
+      state.model = newModel
+      printOk(t.addModelSaved(displayName || model))
+    } catch (err) {
+      printError(err.message)
+    } finally {
+      state.inWizard = false
+    }
+    return 'ok'
+  }
+  if (cmd === '/cd') {
+    if (!arg) { console.log(state.ctx.cwd); return 'ok' }
+    const { resolve } = await import('node:path')
+    const { promises: fsp } = await import('node:fs')
+    const target = resolve(state.ctx.cwd, arg)
+    try {
+      const st = await fsp.stat(target)
+      if (!st.isDirectory()) throw new Error('not a dir')
+      state.ctx.cwd = target
+      // 项目级 skill 与项目级 mcp.json 跟着工作目录走:先重载扩展(会关掉旧目录起的 MCP 子进程),
+      // 再重建首条 system 消息 —— 顺序反了会把旧目录的 skill 清单写进新目录的提示词。
+      await state.extensions?.reload({ cwd: target, locale: state.locale, onWarn: printWarn })
+      state.messages[0] = { role: 'system', content: await buildProjectPrompt({ cwd: target, locale: state.locale, shellDesc: state.shellDesc, extra: extensionSuffix(state) }) }
+      await state.persistSession?.()
+      printOk(t.cdOk(target))
+    } catch {
+      printError(t.cdFail(target))
+    }
+    return 'ok'
+  }
+  printWarn(`未知命令: ${cmd},输入 /help 查看可用命令`)
+  return 'ok'
+}
+
+// ──────────────────────────────────────────────
+// 主入口
+// ──────────────────────────────────────────────
+export async function runAiAgent(argv = []) {
+  const cfg = await config.loadConfig()
+  const locale = cfg.locale || 'zh-CN'
+  const t = makeStrings(locale)
+
+  // g ai --help:智能体专属帮助,不触发主 CLI 的 showHelp
+  if (argv.includes('--help') || argv.includes('-h')) {
+    printSlashHelp(t, locale)
+    return
+  }
+
+  let models = Array.isArray(cfg.models) ? cfg.models : []
+  if (models.length === 0) {
+    // 未配置模型:启动交互式配置向导,参照 g ui 设置里的"添加模型"
+    const setupResult = await runModelSetup({ locale })
+    if (!setupResult) {
+      process.exitCode = 1
+      return
+    }
+    models = setupResult.models
+  }
+
+  // --model=<序号|名称> 选择模型;默认 isDefault,否则第一个
+  let model = models.find(m => m.isDefault) || models[0]
+  const modelArg = argv.find(a => a.startsWith('--model='))
+  if (modelArg) {
+    const q = modelArg.split('=')[1]
+    const idx = Number.parseInt(q, 10)
+    if (Number.isFinite(idx) && idx >= 1 && idx <= models.length) {
+      model = models[idx - 1]
+    } else {
+      const found = models.find(m => modelLabel(m).toLowerCase().includes(String(q).toLowerCase()))
+      if (found) model = found
+      else printWarn(t.unknownModel(q))
+    }
+  }
+
+  const cwd = getCwd()
+  const shellDesc = process.platform === 'win32'
+    ? 'cmd.exe (Windows CMD)'
+    : (process.env.SHELL || '/bin/sh')
+
+  // 加载扩展(Skill 清单 + MCP 服务)。MCP 要起子进程,npx 首次还得下载包,
+  // 所以这一步可能明显停顿 —— 有 MCP 时先提示一句。任何失败都只降级并打印原因。
+  const extensions = await AgentExtensions.load({
+    cwd,
+    locale,
+    onWarn: printWarn,
+    onNotice: (count) => printDim(t.extLoadingMcp(count)),
+  })
+
+  const state = {
+    // 会话持久化:CLI 对话也保存到 ~/.zen-gitsync/agent-sessions/,
+    // 在 Web UI 智能体 tab 中可见(带 CLI 标记)
+    sessionId: genSessionId(),
+    sessionCreatedAt: new Date().toISOString(),
+    messages: [{ role: 'system', content: await buildProjectPrompt({ cwd, locale, shellDesc, extra: extensions.promptSuffix(locale) }) }],
+    ctx: {
+      cwd,
+      onChild: (child) => {
+        state.currentChild = child
+        // 子进程退出后清掉引用,避免 SIGINT 时 kill 一个已结束的进程
+        child.once?.('exit', () => { if (state.currentChild === child) state.currentChild = null })
+      },
+    },
+    models,
+    model,
+    locale,
+    shellDesc,
+    // 单轮工具调用上限:来自全局配置 aiMaxToolIterations(loadConfig 已规范化),
+    // 读不到时由 runAgentTurn 兜底为 200
+    maxToolIterations: cfg.aiMaxToolIterations,
+    // 单轮请求的上下文预算(token 上限/字符换算/条数上限/单条 user 截断线)。
+    // 解析公式与默认值都在 context.js 的 resolveRequestBudget —— CLI 与 Web 共用
+    // 同一份,别在这里另算一套。turn.js 每轮把它透传给 prepareRequestMessages。
+    requestBudget: resolveRequestBudget(cfg.aiMaxRequestTokens),
+    currentChild: null,
+    abortController: null,
+    cancelRequested: false,
+    busy: false,
+    showThinking: true,     // /think 切换:是否回显模型的思考过程
+    thinkingMode: 'full',
+    fullTools: false,
+    lastTurnStats: null,
+    sessionStats: null,
+    pendingImages: [],      // Alt+V / /image 附加的待发送图片 [{path, bytes}]
+    pasting: false,         // 剪贴板读取进行中(防 Alt+V 连打并发)
+    inWizard: false,        // /addmodel 交互式向导进行中:忽略 REPL 的 line 事件
+    rl: null,               // REPL readline 引用(供 /addmodel 复用做 rl.question)
+    extensions,             // 已加载的 Skill + MCP 扩展(见 extensions.js)
+  }
+
+  // 持久化当前会话到磁盘(供 Web UI 读取)
+  let saving = Promise.resolve()
+  async function persistSession() {
+    if (state.messages.length <= 1) return
+    try {
+      const title = autoTitle(state.messages)
+      const sessionId = state.sessionId
+      const snapshot = JSON.parse(JSON.stringify({
+        version: 1,
+        sessionId: state.sessionId,
+        title,
+        source: 'cli',
+        cwd: state.ctx.cwd,
+        model: modelLabel(state.model),
+        createdAt: state.sessionCreatedAt,
+        updatedAt: new Date().toISOString(),
+        messages: repairToolHistory(state.messages),
+        lastTurnStats: state.lastTurnStats,
+        sessionStats: state.sessionStats,
+      }))
+      saving = saving.catch(() => {}).then(() => writeSession(sessionId, snapshot))
+      await saving
+      enforceRetention().catch(() => {})
+    } catch (err) {
+      // 持久化失败不影响 CLI 正常使用
+      printDim(`(会话保存失败: ${err.message})`)
+    }
+  }
+  state.persistSession = persistSession
+
+  // SIGINT 时中止进行中的 LLM 请求 + 正在跑的子命令
+  // (进程退出与全局子进程 drain 由 gitCommit.js 的统一 handler 负责,这里只做快速止血)
+  registerCleanup('aiAgent', async () => {
+    state.cancelRequested = true
+    try { state.abortController?.abort() } catch (_) {}
+    await terminateCommand(state.currentChild)
+    // Ctrl+C 硬退时终端仍停在 bracketed paste 模式 → 之后 shell 里粘贴会带出 200~/201~ 乱码。
+    // 这个开关是终端层的,进程死了没人替它关,必须在这里补一刀。
+    disableBracketedPaste()
+    // MCP 子进程必须显式收掉:它们挂在 stdio 上,不收会让进程无法自然退出
+    await extensions.close()
+    await persistSession()
+  })
+
+  // 扩展摘要:一条 dim 行,没有装扩展时完全静默(extNone 为空串)
+  const extSummary = extensions.skills.length || extensions.tools.length
+    ? t.extLoaded(extensions.skills.length, extensions.tools.length)
+    : t.extNone
+
+  // 单发模式:g ai "帮我改个 bug"(位置参数拼成 prompt,跑一轮就退出)
+  const oneShot = argv.filter(a => !a.startsWith('-')).join(' ').trim()
+  if (oneShot) {
+    console.log(chalk.dim(`[${t.bannerModel}] ${modelLabel(model)} · [${t.bannerCwd}] ${cwd}`))
+    if (extSummary) printDim(extSummary)
+    state.abortController = new AbortController()
+    const stats = await runAgentTurn(state, oneShot, t)
+    if (stats.status === 'completed') printDim(t.oneShotDone)
+    else process.exitCode = stats.status === 'cancelled' ? 130 : 1
+    return
+  }
+
+  // ── 交互模式(REPL)──
+  printBanner({
+    title: 'Zen GitSync — AI Agent',
+    modelLabel: modelLabel(model),
+    baseURL: model.baseURL || '',
+    cwd,
+    modelText: t.bannerModel,
+    cwdText: t.bannerCwd,
+    tip: t.bannerTip,
+  })
+  if (extSummary) printDim(extSummary)
+
+  const rl = readline.createInterface({
+    input: process.stdin,
+    output: process.stdout,
+    // 颜色闭合，避免提示符的样式泄漏到后续输出。
+    prompt: chalk.cyan.bold(t.prompt + ' '),
+    historySize: 200,
+  })
+  state.rl = rl  // 供 /addmodel 等需要 rl.question 的斜杠命令复用
+
+  // 多行粘贴:终端里粘一整段文本时,readline 会把每个 \n 当 Enter —— 一次粘贴变成多次提交,
+  // 第 2 条起还会撞上下面的 busy 分支被丢掉。这里在数据进 readline 之前把整段捕获成占位符,
+  // 提交时再还原(详见 paste.js)。必须在处理 line 事件之前装好。
+  const pasteCapture = installPasteCapture(rl, {
+    formatToken: ({ index, lines }) => t.pasteToken(index, lines),
+  })
+
+  // ask_user pauses the tool loop inside the same readline session. The REPL
+  // line handler is bypassed while this small wizard owns the input line.
+  state.ctx.askUser = ({ question, options = [], allowFreeText = true, multiple = false }) => {
+    const zh = !String(state.locale || '').startsWith('en')
+    const safeOptions = Array.isArray(options) ? options.filter(Boolean).slice(0, 20) : []
+    // 多选只在真的给了选项时才有意义（与 tools.js 的归一化口径一致）
+    const multi = multiple === true && safeOptions.length > 0
+    state.inWizard = true
+    if (safeOptions.length > 0) {
+      console.log(chalk.cyan(`\n${question}`))
+      safeOptions.forEach((option, index) => console.log(`  ${chalk.cyan(String(index + 1))}. ${option}`))
+    } else {
+      console.log(chalk.cyan(`\n${question}`))
+    }
+
+    return new Promise(resolve => {
+      let settled = false
+      const questionController = new AbortController()
+      const parentSignal = state.abortController?.signal
+      const cleanup = () => {
+        rl.removeListener?.('close', onClose)
+        rl.removeListener?.('operationCancel', onCancel)
+        rl.input?.removeListener?.('keypress', onKeypressCancel)
+        parentSignal?.removeEventListener?.('abort', onParentAbort)
+      }
+      const finish = answer => {
+        if (settled) return
+        settled = true
+        cleanup()
+        state.inWizard = false
+        resolve(answer)
+      }
+      const cancel = () => {
+        if (settled) return
+        questionController.abort()
+        finish('Cancelled by user; the question was not answered.')
+      }
+      const onClose = () => cancel()
+      const onCancel = () => {
+        state.cancelRequested = true
+        try { state.abortController?.abort() } catch (_) {}
+        cancel()
+      }
+      const onParentAbort = () => cancel()
+      const onKeypressCancel = (_str, key) => {
+        if (!key?.ctrl || key.name !== 'c') return
+        key.name = 'escape'
+        key.sequence = ''
+        key.ctrl = false
+        key.meta = false
+        key.shift = false
+        onCancel()
+      }
+
+      rl.once('close', onClose)
+      rl.once('operationCancel', onCancel)
+      rl.input?.prependListener?.('keypress', onKeypressCancel)
+      parentSignal?.addEventListener?.('abort', onParentAbort, { once: true })
+
+      const promptAnswer = () => {
+        const hint = safeOptions.length > 0
+          ? (multi
+              ? (zh ? '请输入序号(逗号分隔可多选)' : 'Choose numbers, comma-separated for multiple')
+              : allowFreeText ? (zh ? '请输入序号或直接输入回答' : 'Choose a number or type an answer') : (zh ? '请输入序号' : 'Choose a number'))
+          : (zh ? '请输入回答' : 'Type your answer')
+        rl.question(`${hint}: `, { signal: questionController.signal }, answer => {
+          if (settled) return
+          const trimmed = String(answer || '').trim()
+          if (multi) {
+            // 逗号分隔的序号 → 选项列表(JSON 数组字符串回给模型,与 Web 端口径一致);
+            // 整行不是合法序号串时,允许自由输入的话就把它当作唯一一项
+            const parts = trimmed.split(/[,，]/).map(part => part.trim()).filter(Boolean)
+            const indexes = parts.map(part => Number.parseInt(part, 10))
+            const allIndexes = parts.length > 0 && indexes.every((n, i) => Number.isInteger(n) && String(n) === parts[i] && n >= 1 && n <= safeOptions.length)
+            if (allIndexes) {
+              finish(JSON.stringify(indexes.map(n => safeOptions[n - 1])))
+              return
+            }
+            if (allowFreeText && trimmed) {
+              finish(JSON.stringify([trimmed]))
+              return
+            }
+            console.log(chalk.yellow(zh ? `请输入 1-${safeOptions.length} 之间的序号,多个用逗号分隔` : `Enter numbers from 1 to ${safeOptions.length}, comma-separated`))
+            promptAnswer()
+            return
+          }
+          if (safeOptions.length > 0) {
+            const index = Number.parseInt(trimmed, 10)
+            if (Number.isInteger(index) && index >= 1 && index <= safeOptions.length) {
+              finish(safeOptions[index - 1])
+              return
+            }
+            if (!allowFreeText || !trimmed) {
+              console.log(chalk.yellow(zh ? `请输入 1-${safeOptions.length} 之间的序号` : `Enter a number from 1 to ${safeOptions.length}`))
+              promptAnswer()
+              return
+            }
+          }
+          if (!trimmed) {
+            console.log(chalk.yellow(zh ? '回答不能为空' : 'Answer cannot be empty'))
+            promptAnswer()
+            return
+          }
+          finish(trimmed)
+        })
+      }
+      promptAnswer()
+    })
+  }
+
+  // 输入提示符管理。
+  //
+  // readline 的 _refreshLine() 用 prevRows 追踪“提示符占了几行”,
+  // 下次刷新时 moveCursor(0, -prevRows) 上移光标。但模型输出直接写
+  // stdout,readline 不知道这些行,导致 prevRows 过期 → 光标错位。
+  //
+  // 修复:每次画提示符前,先写 \r\n 确保光标在新行第 0 列,
+  // 然后重置 prevRows=0 告诉 readline “当前行就是提示符行”。
+  // 这样 _refreshLine() 的 clearScreenDown() 只清当前空行,不影响模型输出。
+
+  const showPrompt = () => {
+    if (rl.closed) return
+    // 双保险:确保 stdin 处于 flowing 模式(ora 的 stdin-discarder 可能在
+    // stop() 时 pause 了 stdin,即使 discardStdin:false 已设,仍防止其他
+    // 库意外 pause)
+    if (process.stdin.isTTY && !process.stdin.readableFlowing) {
+      process.stdin.resume()
+    }
+    // bracketed paste 是终端层的开关,跑过的外部命令(全屏 TUI 之类)退出时可能把它关掉,
+    // 之后粘贴又会退回"每个 \n 一次提交"。它是幂等的,每次画提示符时重申一次。
+    enableBracketedPaste()
+    if (process.stdout.isTTY) {
+      // 确保光标在新行第 0 列(\r 回行首,\n 换行)
+      process.stdout.write('\r\n')
+      // 重置 readline 内部光标追踪
+      rl.prevRows = 0
+    }
+    rl.prompt()
+  }
+
+  // 安全版:防止 finally 块里的异常变成 unhandledRejection 杀掉会话
+  const safeShowPrompt = () => {
+    try {
+      showPrompt()
+    } catch (_) {
+      try { rl.prompt() } catch { /* 给予宽容 */ }
+    }
+  }
+
+  // 仅重画提示符,不写 \r\n(用于空输入、busy 等不需要换行的场景)
+  const safeRefreshPrompt = () => {
+    try {
+      if (rl.closed) return
+      enableBracketedPaste()
+      rl.prevRows = 0
+      rl.prompt()
+    } catch (_) {
+      try { rl.prompt() } catch { /* noop */ }
+    }
+  }
+
+  // 在提示符上方插一行通知(粘贴进度等)。
+  const notifyAbovePrompt = (text) => {
+    console.log(text)
+    if (!state.busy) safeShowPrompt()
+  }
+
+  // ── 斜杠命令即时提示(支持 ↑↓ 切换 / Tab、Enter 补全) ──
+  // 在提示符下方浮现匹配到的命令,随输入过滤;不移动光标(用 DECSC/DECRC 保存-恢复)。
+  // slashHintRows 记录当前提示占了几行,下次刷新/提交时据此擦除,避免残影。
+  // slashHintBaseLine:本次 hint 显示时"输入行的基线值" — ↑↓ 时 readline 已经把
+  //   rl.line 改成历史上一行,我们用基线回滚 + _refreshLine() 重绘,让用户视觉上看
+  //   不到"按 ↑ 输入行变了",只看到 hint 高亮跳了一格。
+  let slashHintRows = 0
+  let slashHintBaseLine = ''
+  let slashHintMatches = []
+  let slashHintIndex = 0
+  const eraseSlashHint = () => {
+    if (!process.stdout.isTTY || slashHintRows === 0) return
+    // 保存光标 → 下移到提示区逐行清空 → 恢复光标
+    let seq = '\x1b7'
+    for (let i = 0; i < slashHintRows; i++) seq += '\x1b[1B\x1b[2K'
+    seq += '\x1b8'
+    process.stdout.write(seq)
+    slashHintRows = 0
+  }
+  const drawSlashHint = () => {
+    const body = renderSlashHintBody(slashHintMatches, slashHintIndex, state.locale)
+    const hadVisibleHint = slashHintRows > 0
+    eraseSlashHint()
+    if (!body) return
+    const rows = body.split('\n')
+    if (!hadVisibleHint) {
+      // CSI B(光标下移)到达终端底部后只会停在最后一行,不会自动滚屏,
+      // 所以候选项会被裁掉。第一次显示提示时先用换行真实预留空间,
+      // 再上移相同行数回到输入行；无论中途是否发生滚屏,都能回到原内容所在行。
+      let reserve = ''
+      for (let i = 0; i < rows.length; i++) reserve += '\n\x1b[2K'
+      reserve += `\x1b[${rows.length}A`
+      process.stdout.write(reserve)
+    }
+    // 保存光标,逐行下移打印(每行先清行防止与旧内容叠字),最后恢复光标回输入行
+    let seq = '\x1b7'
+    for (const r of rows) seq += '\x1b[1B\x1b[2K\r' + r
+    seq += '\x1b8'
+    process.stdout.write(seq)
+    slashHintRows = rows.length
+  }
+  // 重置 hint 上下文:重新计算匹配 + 重画。输入字符变化时调用。
+  const refreshSlashHint = () => {
+    if (!process.stdout.isTTY || state.busy || state.inWizard) {
+      eraseSlashHint()
+      slashHintMatches = []
+      slashHintIndex = 0
+      slashHintBaseLine = ''
+      return
+    }
+    slashHintBaseLine = rl.line
+    slashHintMatches = filterSlashCommands(rl.line, state.locale)
+    // 匹配列表缩短时夹紧 index(增删交替时不会越界)
+    slashHintIndex = slashHintIndex < slashHintMatches.length ? slashHintIndex : 0
+    drawSlashHint()
+  }
+  // ↑↓ / Shift+Tab:readline 已经把 rl.line 改成历史行或没动 → 强制恢复基线并重绘输入行
+  const restoreInputLineToBaseline = () => {
+    if (rl.line === slashHintBaseLine) return
+    rl.line = slashHintBaseLine
+    rl.cursor = slashHintBaseLine.length
+    if (typeof rl._refreshLine === 'function') rl._refreshLine()
+  }
+  // Tab / Enter:把 rl.line 替换为补全文本 + 清掉 hint 状态。
+  // Enter 会先被 readline 提交,因此由下方 line handler 在处理命令前调用本函数。
+  const applyHintCompletion = (newLine) => {
+    rl.line = newLine
+    rl.cursor = newLine.length
+    eraseSlashHint()
+    slashHintMatches = []
+    slashHintIndex = 0
+    slashHintRows = 0
+    slashHintBaseLine = newLine
+    if (typeof rl._refreshLine === 'function') rl._refreshLine()
+  }
+
+  const neutralizeKeyForReadline = (key) => {
+    key.name = 'escape'
+    key.sequence = ''
+    key.ctrl = false
+    key.meta = false
+    key.shift = false
+  }
+
+  // readline 自己会优先处理 Enter/Ctrl+C。用 prependListener 在它之前完成补全
+  // 或取消当前操作,再把本次按键变成 readline 会忽略的 Esc。
+  const interceptInteractiveControlKeys = (_ch, key) => {
+    if (!key) return
+    if (key.ctrl && key.name === 'c') {
+      if (state.inWizard) {
+        neutralizeKeyForReadline(key)
+        rl.emit('operationCancel')
+        return
+      }
+      if (state.busy) {
+        neutralizeKeyForReadline(key)
+        state.cancelRequested = true
+        try { state.abortController?.abort() } catch (_) {}
+        void terminateCommand(state.currentChild)
+        return
+      }
+      if (slashHintRows > 0) {
+        neutralizeKeyForReadline(key)
+        applyHintCompletion('')
+        return
+      }
+      return
+    }
+    if (key.name !== 'return' && key.name !== 'enter') return
+    if (slashHintRows === 0 || slashHintMatches.length === 0 || state.busy || state.inWizard) return
+    const selected = slashHintMatches[slashHintIndex]
+    if (!selected) return
+    applyHintCompletion(selected.cmd + ' ')
+    neutralizeKeyForReadline(key)
+  }
+  rl.input.prependListener('keypress', interceptInteractiveControlKeys)
+
+  // Alt+V 粘贴剪贴板图片(node 把 ESC+v 解析为 meta+v;部分终端不发 Alt,可用 /image 兜底)
+  const pasteFromClipboard = async () => {
+    if (state.pasting) return
+    state.pasting = true
+    try {
+      notifyAbovePrompt(chalk.dim(t.imagePasting))
+      const img = await readClipboardImage()
+      if (!img) {
+        notifyAbovePrompt(chalk.yellow(t.imageEmpty))
+      } else {
+        state.pendingImages.push(img)
+        notifyAbovePrompt(chalk.green(t.imageAttached(state.pendingImages.length, formatBytes(img.bytes))))
+      }
+    } catch (err) {
+      // 不能让粘贴异常变成 unhandledRejection 杀掉会话
+      printError(`粘贴图片失败: ${err.message}`)
+    } finally {
+      state.pasting = false
+    }
+  }
+  rl.input.on('keypress', (ch, key) => {
+    if (!key) return
+
+    // ── 斜杠命令提示专用键盘:↑↓/Tab/Enter/Shift+Tab/Esc ──
+    // 仅在 hint 已显示且有匹配项时拦截,readline 的 default 行为(历史浏览等)
+    // 我们在 handler 里事后回滚 rl.line 来抵消
+    const action = parseKeyForSlashHint(key)
+    const hintActive = action === 'prev' || action === 'next' || action === 'complete' || action === 'cancel'
+      ? (slashHintRows > 0 && slashHintMatches.length > 0 && !state.busy && !state.inWizard)
+      : false
+
+    if (hintActive) {
+      if (action === 'prev' || action === 'next') {
+        const n = slashHintMatches.length
+        slashHintIndex = action === 'prev'
+          ? (slashHintIndex - 1 + n) % n
+          : (slashHintIndex + 1) % n
+        restoreInputLineToBaseline()
+        drawSlashHint()
+        return   // 不再走 default update + Alt+V 分支
+      }
+      if (action === 'complete') {
+        // Tab:把输入行替换为选中命令 + 末尾空格(便于继续打参数),
+        // 光标跳到末尾;之后不再触发 hint(已经精确选了某个命令)
+        const sel = slashHintMatches[slashHintIndex]
+        const newLine = sel.cmd + ' '
+        applyHintCompletion(newLine)
+        return
+      }
+      if (action === 'cancel') {
+        // Esc:只清 hint,不动输入行
+        eraseSlashHint()
+        slashHintMatches = []
+        slashHintIndex = 0
+        return
+      }
+    }
+
+    // Enter 走默认 readline 提交,line 事件里把选中项恢复到输入框;
+    // 其他普通按键:重算匹配 + 重画 hint
+    if (!(key.name === 'return' || key.name === 'enter')) {
+      refreshSlashHint()
+    }
+
+    // Alt+V 粘贴剪贴板图片(node 把 ESC+v 解析为 meta+v;部分终端不发 Alt,可用 /image 兜底)
+    const isAltV = (key.meta && key.name === 'v')
+      || (typeof key.sequence === 'string' && key.sequence.length === 2
+          && key.sequence.charCodeAt(0) === 27 && key.sequence[1] === 'v')
+    if (isAltV) pasteFromClipboard()
+  })
+
+  // 初始提示符:banner 后直接画(不需要 \r\n)
+  rl.prevRows = 0
+  rl.prompt()
+
+  rl.on('line', async (line) => {
+    // 提交时光标已落到提示符下一行(即提示区起点),直接清到屏幕末尾擦掉整块提示,
+    // 比 DECSC/DECRC 相对定位更稳(此时光标不在输入行)
+    if (process.stdout.isTTY && slashHintRows > 0) {
+      process.stdout.write('\x1b[0J')
+      slashHintRows = 0
+    }
+    // /addmodel 等交互式向导进行中时,用户的回答由向导自身的 rl.question 处理,
+    // 不应进入 REPL 的正常输入流程(否则会把向导的答案当成命令/消息发给模型)
+    if (state.inWizard) return
+    // 把输入行里的粘贴占位符还原成原文(带真实换行)—— 在判断是不是斜杠命令之前,
+    // 这样 `请写进 .env [粘贴 #1]` 这种混排也能正确展开
+    const pasted = pasteCapture.expand(line)
+    if (pasted.stale.length > 0) printWarn(t.pasteExpired(pasted.stale.length))
+    const input = pasted.text.trim()
+
+    if (input.startsWith('/')) {
+      const r = await handleSlashCommand(state, input, t)
+      if (r === 'exit') { rl.close(); return }
+      safeRefreshPrompt()
+      return
+    }
+
+    if (!input) { safeRefreshPrompt(); return }
+    if (state.busy) {
+      printDim(t.busy)
+      safeRefreshPrompt()
+      return
+    }
+
+    // 取出待发送图片(取出即清空队列,用户每条消息独立决定带不带图)
+    const images = state.pendingImages.splice(0)
+    if (images.length > 0) printDim(t.imageSending(images.length))
+
+    // 用了粘贴的这条消息:输入行里只有占位符,这里把真正发出去的内容亮一遍,
+    // 否则回看终端记录只剩一个 [粘贴 #N],等于把内容藏起来了。
+    // 只在 TTY 里回显 —— 管道/CI 下 readline 本来就不回显输入行(没有占位符这回事),
+    // 再补一块只会粘在提示符后面变成半行。
+    if (pasted.used > 0 && process.stdout.isTTY) process.stdout.write(renderPasteEcho(input, { locale: state.locale }))
+
+    state.busy = true
+    state.cancelRequested = false
+    state.abortController = new AbortController()
+    try {
+      await runAgentTurn(state, input, t, images)
+    } catch (err) {
+      if (state.cancelRequested) printDim(t.operationCancelled)
+      else printError(err.message)
+    } finally {
+      state.busy = false
+      state.cancelRequested = false
+      state.abortController = null
+      safeShowPrompt()
+    }
+  })
+
+  rl.on('close', () => {
+    if (state.busy) {
+      state.cancelRequested = true
+      state.abortController?.abort()
+      void terminateCommand(state.currentChild)
+    }
+    rl.input.removeListener('keypress', interceptInteractiveControlKeys)
+    pasteCapture.dispose()   // 卸载数据改写 + 把终端从 bracketed paste 模式里放出来
+    printWarn('\n' + t.bye)
+  })
+
+  // 全局安全网:任何未捕获的 Promise rejection / 异常都不应该静默杀掉
+  // 整个 REPL 会话。这里记录到 stderr 但不退出,让 readline 继续运行。
+  const rejectionHandler = (reason) => {
+    const msg = reason instanceof Error ? reason.message : String(reason)
+    printError(`内部错误(已捕获,会话继续): ${msg}`)
+    safeShowPrompt()
+  }
+  process.on('unhandledRejection', rejectionHandler)
+
+  // 等 readline 关闭后返回(进程随后自然退出)
+  await new Promise((resolve) => rl.on('close', resolve))
+
+  // 清理:移除本次会话注册的全局处理器
+  process.removeListener('unhandledRejection', rejectionHandler)
+}
+
+export default { runAiAgent }