{
  "generatedBy": "scripts/write-cli-startup-metadata.ts",
  "generatorSignature": "a0500be4a86b858a5c269fe9edb43d43fc4b8379",
  "channelOptions": [
    "telegram",
    "whatsapp",
    "discord",
    "irc",
    "googlechat",
    "slack",
    "signal",
    "imessage",
    "feishu",
    "nostr",
    "msteams",
    "mattermost",
    "nextcloud-talk",
    "matrix",
    "line",
    "zalo",
    "clickclack",
    "zalouser",
    "sms",
    "synology-chat",
    "tlon",
    "qa-channel",
    "qqbot",
    "twitch"
  ],
  "channelCatalogSignature": "08788cfc2abd2ba6b181428ca8312355bcb6fb3b",
  "rootHelpBundleSignature": "95e6f39457133d17071a5429c25b17f41bac44a9",
  "browserHelpSourceSignature": "b334886e946e6818fe0531c272b6328fbe333974",
  "secretsHelpSourceSignature": "5c3b7160cf3b11fedef896e9252f47639476787e",
  "nodesHelpSourceSignature": "1cc420a0a8312f94109b726c667e4d513b17135f",
  "subcommandHelpSourceSignature": "2d80d1b0777c5d5242c38f2d0b911f903e4e9226",
  "browserHelpText": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw browser [options] [command]\n\nManage OpenClaw's dedicated browser (Chrome/Chromium)\n\n选项：\n  --browser-profile <name>  Browser profile name (default from config)\n  --expect-final            Wait for final response (agent) (default: false)\n  -h, --help                显示命令帮助\n  --json                    Output machine-readable JSON (default: false)\n  --timeout <ms>            Timeout in ms (default: \"30000\")\n  --token <token>           Gateway token (if required)\n  --url <url>               Gateway WebSocket URL (defaults to\n                            gateway.remote.url when configured)\n\n命令：\n  click                     Click an element by ref from snapshot\n  click-coords              Click viewport coordinates\n  close                     Close a tab (tab reference optional)\n  console                   Get recent console messages\n  cookies                   Read/write cookies\n  create-profile            Create a new browser profile\n  delete-profile            Delete a browser profile\n  dialog                    Arm the next modal dialog (alert/confirm/prompt)\n  doctor                    Check browser plugin readiness\n  download                  Click a ref and save the resulting download\n  drag                      Drag from one ref to another\n  errors                    Get recent page errors\n  evaluate                  Evaluate a function against the page or a ref\n  fill                      Fill a form with JSON field descriptors\n  focus                     Focus a tab by tab reference\n  highlight                 Highlight an element by ref\n  hover                     Hover an element by ai ref\n  navigate                  Navigate the current tab to a URL\n  open                      Open a URL in a new tab\n  pdf                       Save page as PDF\n  press                     Press a key\n  profiles                  List all browser profiles\n  requests                  Get recent network requests (best-effort)\n  reset-profile             Reset browser profile (moves it to Trash)\n  resize                    Resize the viewport\n  responsebody              Wait for a network response and return its body\n  screenshot                Capture a screenshot (prints the saved path)\n  scrollintoview            Scroll an element into view by ref from snapshot\n  select                    Select option(s) in a select element\n  set                       Browser environment settings\n  snapshot                  Capture a snapshot (default: ai; aria is the\n                            accessibility tree)\n  start                     Start the browser (no-op if already running)\n  status                    Show browser status\n  stop                      Stop the browser (best-effort)\n  storage                   Read/write localStorage/sessionStorage\n  tab                       Tab shortcuts (index-based)\n  tabs                      List open tabs\n  trace                     Record a Playwright trace\n  type                      Type into an element by ref from snapshot\n  upload                    Arm file upload for the next file chooser\n  wait                      Wait for time, selector, URL, load state, or JS\n                            conditions\n  waitfordownload           Wait for the next download (and save it)\n\nExamples:\n  openclaw browser status\n  openclaw browser start\n  openclaw browser start --headless\n  openclaw browser stop\n  openclaw browser tabs\n  openclaw browser open https://example.com\n  openclaw browser focus abcd1234\n  openclaw browser close abcd1234\n  openclaw browser screenshot\n  openclaw browser screenshot --full-page\n  openclaw browser screenshot --ref 12\n  openclaw browser snapshot\n  openclaw browser snapshot --format aria --limit 200\n  openclaw browser snapshot --efficient\n  openclaw browser snapshot --labels\n  openclaw browser navigate https://example.com\n  openclaw browser resize 1280 720\n  openclaw browser click 12 --double\n  openclaw browser click-coords 120 340\n  openclaw browser type 23 \"hello\" --submit\n  openclaw browser press Enter\n  openclaw browser hover 44\n  openclaw browser drag 10 11\n  openclaw browser select 9 OptionA OptionB\n  openclaw browser upload /tmp/openclaw/uploads/file.pdf\n  openclaw browser upload media://inbound/file.pdf\n  openclaw browser fill --fields '[{\"ref\":\"1\",\"value\":\"Ada\"}]'\n  openclaw browser dialog --accept\n  openclaw browser wait --text \"Done\"\n  openclaw browser evaluate --fn '(el) => el.textContent' --ref 7\n  openclaw browser console --level error\n  openclaw browser pdf\n\nDocs: https://docs.openclaw.ai/cli/browser\n\n",
  "secretsHelpText": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw secrets [options] [command]\n\nSecrets runtime controls\n\n选项：\n  -h, --help  显示命令帮助\n\n命令：\n  apply       Apply a previously generated secrets plan\n  audit       Audit plaintext secrets, unresolved refs, and precedence drift\n  configure   Interactive secrets helper (provider setup + SecretRef mapping +\n              preflight)\n  help        显示命令帮助\n  reload      Re-resolve secret references and atomically swap runtime snapshot\n\nDocs: https://docs.openclaw.ai/gateway/security\n\n",
  "nodesHelpText": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw nodes [options] [command]\n\nManage gateway-owned nodes (pairing, status, invoke, and media)\n\n选项：\n  -h, --help  显示命令帮助\n\n命令：\n  approve     Approve a pending pairing request\n  camera      Capture camera media from a paired node\n  canvas      Capture or render canvas content from a paired node\n  describe    Describe a node (capabilities + supported invoke commands)\n  help        显示命令帮助\n  invoke      Invoke a command on a paired node\n  list        List pending and paired nodes\n  location    Fetch location from a paired node\n  notify      Send a local notification on a node (mac only)\n  pending     List pending pairing requests\n  push        Send an APNs test push to an iOS node\n  reject      Reject a pending pairing request\n  remove      Remove a paired node entry\n  rename      Rename a paired node (display name override)\n  screen      Capture screen recordings from a paired node\n  status      List known nodes with connection status and capabilities\n\nExamples:\n  openclaw nodes status\n    List known nodes with live status.\n  openclaw nodes pairing pending\n    Show pending node pairing requests.\n  openclaw nodes remove --node <id|name|ip>\n    Remove a stale paired node entry.\n  openclaw nodes invoke --node <id> --command system.which --params '{\"name\":\"uname\"}'\n    Invoke a node command directly.\n  openclaw nodes camera snap --node <id>\n    Capture a photo from a node camera.\n\nDocs: https://docs.openclaw.ai/cli/nodes\n\n",
  "subcommandHelpText": {
    "doctor": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw doctor [options]\n\nHealth checks + quick fixes for the gateway and channels\n\n选项：\n  --allow-exec                Allow doctor to execute exec SecretRefs while\n                              verifying configured secrets (default: false)\n  --deep                      Scan system services for extra gateway installs\n                              (default: false)\n  --fix                       Apply recommended repairs (alias for --repair)\n                              (default: false)\n  --force                     Apply aggressive repairs (overwrites custom\n                              service config) (default: false)\n  --generate-gateway-token    Generate and configure a gateway token (default:\n                              false)\n  -h, --help                  显示命令帮助\n  --json                      With --lint or --post-upgrade: emit\n                              machine-readable JSON output (default: false)\n  --lint                      Run read-only health checks and report findings\n                              (default: false)\n  --no-workspace-suggestions  Disable workspace memory system suggestions\n  --non-interactive           Run without prompts (safe migrations only)\n                              (default: false)\n  --only <id>                 With --lint: run only the specified check id\n                              (repeatable) (default: [])\n  --post-upgrade              Emit plugin-compat findings only (machine-readable\n                              with --json) (default: false)\n  --repair                    Apply recommended repairs without prompting\n                              (default: false)\n  --severity-min <level>      With --lint: drop findings below this severity\n                              (info|warning|error)\n  --skip <id>                 With --lint: skip a specific check id (repeatable)\n                              (default: [])\n  --yes                       Accept defaults without prompting (default: false)\n\nDocs: https://docs.openclaw.ai/cli/doctor\n\n",
    "gateway": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw gateway [options] [command]\n\n运行、检查和查询 WebSocket 网关\n\n选项：\n  --allow-unconfigured       允许网关启动而不强制配置中的 gateway.mode=local（不修复配置） (default:\n                             false)\n  --auth <mode>              网关认证模式（\"none\"|\"token\"|\"password\"|\"trusted-proxy\"）\n  --bind <mode>              绑定模式（\"loopback\"|\"lan\"|\"tailnet\"|\"auto\"|\"custom\"）。默认为配置\n                             gateway.bind（或 loopback）。\n  --claude-cli-logs          已弃用，--cli-backend-logs 的别名 (default: false)\n  --cli-backend-logs         仅在控制台显示 CLI 后端日志（包括 stdout/stderr） (default: false)\n  --compact                  \"--ws-log compact\" 的别名 (default: false)\n  --dev                      如缺失则创建开发配置和工作区（不含 BOOTSTRAP.md） (default: false)\n  --force                    启动前终止目标端口上的任何现有监听进程 (default: false)\n  -h, --help                 显示命令帮助\n  --password <password>      auth mode=password 时的密码\n  --password-file <path>     从文件读取网关密码\n  --port <port>              网关 WebSocket 端口\n  --raw-stream               将原始模型流事件记录到 jsonl (default: false)\n  --raw-stream-path <path>   原始流 jsonl 路径\n  --reset                    重置开发配置、凭证、会话和工作区（需要 --dev） (default: false)\n  --tailscale <mode>         Tailscale 暴露模式（\"off\"|\"serve\"|\"funnel\"）\n  --tailscale-reset-on-exit  关闭时重置 Tailscale serve/funnel 配置 (default: false)\n  --token <token>            connect.params.auth.token 中要求的共享令牌（默认：环境变量\n                             OPENCLAW_GATEWAY_TOKEN 如已设置）\n  --verbose                  向 stdout/stderr 输出详细日志 (default: false)\n  --ws-log <style>           WebSocket 日志样式（\"auto\"|\"full\"|\"compact\"） (default:\n                             \"auto\")\n\n命令：\n  call                       调用网关方法\n  diagnostics                Export local support diagnostics\n  discover                   通过 Bonjour 发现网关（本地 + 已配置时的广域）\n  health                     获取网关健康状态\n  install                    安装网关服务（launchd/systemd/schtasks）\n  probe                      显示网关可达性、认证能力和读探测摘要（本地 + 远程）\n  restart                    重启网关服务（launchd/systemd/schtasks）\n  run                        在前台运行 WebSocket 网关\n  stability                  获取不含载荷的网关稳定性诊断\n  start                      启动网关服务（launchd/systemd/schtasks）\n  status                     显示网关服务状态并探测连通性/能力\n  stop                       停止网关服务（launchd/systemd/schtasks）\n  uninstall                  卸载网关服务（launchd/systemd/schtasks）\n  usage-cost                 从会话日志中获取用量成本摘要\n\n示例：\n  openclaw gateway run\n    在前台运行网关。\n  openclaw gateway status\n    显示服务状态以及连通性/能力。\n  openclaw gateway discover\n    查找本地和广域网关信标。\n  openclaw gateway stability\n    显示最近的稳定性诊断。\n  openclaw gateway call health\n    直接调用网关 RPC 方法。\n\n文档： https://docs.openclaw.ai/cli/gateway\n\n",
    "models": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw models [options] [command]\n\n模型发现、扫描和配置\n\n选项：\n  --agent <id>     要检查的代理 ID (覆盖 OPENCLAW_AGENT_DIR)\n  -h, --help       显示命令帮助\n  --status-json    输出 JSON (`models status --json` 的别名) (default: false)\n  --status-plain   纯文本输出 (`models status --plain` 的别名) (default: false)\n\n命令：\n  aliases          管理模型别名\n  auth             管理模型认证配置\n  fallbacks        管理模型备选列表\n  image-fallbacks  管理图像模型备选列表\n  list             列出模型 (默认已配置的)\n  scan             扫描 OpenRouter 免费模型以获取工具+图像支持\n  set              设置默认模型\n  set-image        设置图像模型\n  status           显示已配置的模型状态\n\nDocs: https://docs.openclaw.ai/cli/models\n\n",
    "plugins": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw plugins [options] [command]\n\nManage OpenClaw plugins and extensions\n\n选项：\n  -h, --help   显示命令帮助\n\n命令：\n  build        Generate simple tool plugin metadata\n  disable      Disable a plugin in config\n  doctor       Report plugin load issues\n  enable       Enable a plugin in config\n  init         Create a simple tool plugin project\n  inspect      Inspect plugin details\n  install      Install a plugin or hook pack (path, archive, npm spec, git repo,\n               clawhub:package, or marketplace entry)\n  list         List discovered plugins\n  marketplace  Inspect Claude-compatible plugin marketplaces\n  registry     Inspect or rebuild the persisted plugin registry\n  search       Search ClawHub plugin packages\n  uninstall    Uninstall a plugin\n  update       Update installed plugins and tracked hook packs\n  validate     Validate simple tool plugin metadata\n\nDocs: https://docs.openclaw.ai/cli/plugins\n\n"
  },
  "rootHelpText": "\nOpenClaw 中文版 2026.6.5 (5181e4f) — 所有聊天，一个 OpenClaw 搞定。\n\n用法： openclaw [options] [command]\n\n选项：\n  --container <name>   在正在运行的 Podman/Docker 容器 <name> 内运行 CLI（默认：环境变量\n                       OPENCLAW_CONTAINER）\n  --dev                开发配置：在 ~/.openclaw-dev 下隔离状态，默认网关端口 19001，并移动派生端口（浏览器/画布）\n  -h, --help           显示命令帮助\n  --log-level <level>  全局日志级别覆盖，适用于文件和控制台\n                       (silent|fatal|error|warn|info|debug|trace)\n  --no-color           禁用 ANSI 颜色\n  --profile <name>     使用命名配置文件（在 ~/.openclaw-<name> 下隔离状态和配置）\n  -V, --version        输出版本号\n\n命令：\n  提示：带 * 后缀的命令有子命令。运行 <command> --help 查看详情。\n  acp *                运行和管理基于 ACP 的代码代理\n  agent                通过网关运行一次代理\n  agents *             管理隔离的代理（工作区、认证、路由）\n  approvals *          管理执行审批（网关或节点主机）\n  backup *             创建和验证 OpenClaw 状态的本地备份归档\n  capability *         运行提供商能力命令（备用别名：infer）\n  channels *           添加、删除、登录和查看聊天渠道\n  chat                 打开本地终端界面（tui --local 的别名）\n  clawbot *            旧版 clawbot 命令别名\n  commitments *        列出并管理推断出的后续跟进事项\n  completion           生成 Shell 补全脚本\n  config *             非交互式配置工具（get/set/unset/file/validate）。默认：启动引导式设置。\n  configure            交互式配置凭证、渠道、网关和代理默认设置\n  crestodian           打开交互式设置和修复助手\n  cron *               调度和查看网关后台任务\n  daemon *             管理网关服务（旧版别名）\n  dashboard            使用当前令牌打开控制面板\n  devices *            设备配对和令牌管理\n  directory *          查询支持渠道的联系人和群组 ID\n  dns *                DNS 工具 - 广域发现（Tailscale + CoreDNS）\n  docs                 搜索 OpenClaw 在线文档\n  doctor               诊断并修复配置、网关、插件和渠道问题\n  exec-policy *        查看或同步请求的执行策略与主机审批\n  gateway *            运行、检查和查询 OpenClaw 网关\n  health               从运行中的网关获取详细健康状态\n  help                 显示命令帮助\n  hooks *              管理内部代理 Hooks\n  infer *              运行提供商支持的模型、媒体、搜索和嵌入命令\n  logs                 本地或通过 RPC 查看网关日志\n  mcp *                管理 OpenClaw MCP 配置和渠道桥接\n  memory *             搜索、查看并重建记忆文件索引\n  message *            发送、读取和管理渠道消息\n  migrate *            从其他代理系统导入状态\n  models *             列出、扫描和设置模型提供商\n  node *               运行和管理无头节点主机服务\n  nodes *              通过网关配对节点并运行节点主机命令\n  onboard              交互式引导配置网关、工作区和技能\n  pairing *            安全私信配对（审批入站请求）\n  plugins *            安装、启用、禁用和查看插件\n  proxy *              运行 OpenClaw 调试代理并查看捕获的流量\n  qr                   生成移动端配对二维码/设置码\n  reset                重置本地配置/状态（保留 CLI）\n  sandbox *            管理沙箱容器以隔离代理\n  secrets *            审计、应用和重载基于 SecretRef 的凭证\n  security *           安全工具和本地配置审计\n  sessions *           列出已存储的对话会话\n  setup                初始化本地配置和代理工作区\n  skills *             列出、查看和安装代理技能\n  status               显示网关、渠道、模型和最近会话状态\n  system *             系统事件、心跳和在线状态\n  tasks *              查看持久化后台任务和流程\n  terminal             打开本地终端界面（tui --local 的别名）\n  transcripts *        查看已存储的会话记录\n  tui                  打开连接到网关的终端界面\n  uninstall            卸载网关服务和本地数据（保留 CLI）\n  update *             更新 OpenClaw 并检查更新渠道状态\n  webhooks *           Webhook 工具与集成\n\n示例：\n  openclaw onboard\n    引导设置本地网关、工作区、认证和渠道。\n  openclaw setup\n    创建基础配置、工作区和会话目录。\n  openclaw configure\n    调整模型、网关、渠道、插件、技能和健康检查。\n  openclaw status\n    查看网关、渠道、模型和最近会话状态。\n  openclaw doctor --fix\n    修复常见的配置、服务、插件和渠道问题。\n  openclaw channels add\n    在引导提示下添加或更新聊天渠道账号。\n  openclaw channels status\n    查看已连接的消息账号和登录状态。\n  openclaw --dev gateway\n    在 ws://127.0.0.1:19001 上运行开发网关（隔离状态/配置）。\n  openclaw gateway run --force\n    启动网关并替换占用其端口的进程。\n  openclaw models status\n    运行代理前查看模型/提供商认证健康状态。\n  openclaw plugins list\n    查看已启用、已禁用和已安装的插件。\n  openclaw agent --to +15555550123 --message \"Run summary\" --deliver\n    通过网关运行一次代理轮次，并可选交付回复。\n  openclaw message send --channel telegram --target @mychat --message \"Hi\"\n    通过你的 Telegram 机器人发送。\n\n文档： https://docs.openclaw.ai/cli\n汉化版： https://openclaw.qt.cool/\n\n"
}