Menu
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/GeorgeDoors888/GB-Power-Market-JJ --skill zz-chatThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Showing SKILL.md
Based on SOC occupation classification
| name | zz-chat |
| description | 爪爪/钳钳 — 手机直连 OpenClaw 聊天。触发词：钳钳、爪爪、钱钱、龙虾、聊天机器人、创建爪爪、安装钳钳、打开钳钳、zz-chat。 |
| metadata | {"clawdbot":{"emoji":"🦞"}} |
OC 聊天: 手机 ←WebSocket→ 用户的Worker ←→ bridge ←→ OpenClaw CLI
朋友互聊: 手机A ←WebSocket→ WorkerA ──HTTP──→ WorkerB ←WebSocket→ 手机B
注册发现: Worker ──注册──→ 中心Worker（ai0000.cn）←──查询── Worker
每个用户独立部署 Worker，用自己 Cloudflare 额度，互不影响。
用户说"部署爪爪"或类似指令时，按以下步骤自动执行：
安装 wrangler（如果没装）：
export PATH="/usr/local/bin:$PATH"
which wrangler 2>/dev/null || npm install -g wrangler 2>&1
检查登录状态，未登录则弹出浏览器让用户授权：
wrangler auth whoami 2>&1
如果显示"Not logged in"，执行登录（会自动弹浏览器，用户点同意即可）：
wrangler login 2>&1
# 等用户在浏览器中点 "Allow" 授权
准备用户版 Worker 文件：
cd ~/.openclaw/workspace/openchat/simple
# 用用户版文件覆盖中心版
cp worker-user.js worker-user-deploy.js
cp wrangler-user.toml wrangler-user-deploy.toml
修改 wrangler-user-deploy.toml 中的 name，改成用户喜欢的名字（如 zz-chat-xxx）：
sed -i '' "s/name = \"zz-proxy\"/name = \"zz-chat-$(whoami)\"/" wrangler-user-deploy.toml
部署：
cd ~/.openclaw/workspace/openchat/simple
export PATH="/usr/local/bin:$PATH"
wrangler deploy -c wrangler-user-deploy.toml 2>&1
记录 Worker URL：
WORKER_URL=$(wrangler deploy -c wrangler-user-deploy.toml 2>&1 | grep -o 'https://[^ ]*workers.dev' | head -1)
mkdir -p ~/.zz
echo "$WORKER_URL" > ~/.zz/worker_url
echo "Worker URL: $WORKER_URL"
WORKER_URL=$(cat ~/.zz/worker_url)
ZZ_ID=$(curl -s "$WORKER_URL/register" | python3 -c "import sys,json; print(json.load(sys.stdin)['id'])")
echo "$ZZ_ID" > ~/.zz/id
echo "编号: $ZZ_ID"
先杀掉旧进程：
pkill -f "bridge.py" 2>/dev/null
pkill -f "watchdog.py" 2>/dev/null
sleep 1
复制桥接脚本：
BRIDGE_DIR=~/.openclaw/workspace/openchat/bridge
mkdir -p "$BRIDGE_DIR"
cp ~/.openclaw/skills/zz-chat/bridge.py "$BRIDGE_DIR/"
cp ~/.openclaw/skills/zz-chat/watchdog.py "$BRIDGE_DIR/"
启动看门狗（自动拉起 bridge）：
WORKER_URL=$(cat ~/.zz/worker_url)
ZZ_ID=$(cat ~/.zz/id)
nohup python3 -u "$BRIDGE_DIR/watchdog.py" >> "$BRIDGE_DIR/watchdog.log" 2>&1 &
sleep 5
# 检查是否启动成功
ps aux | grep -E "(bridge|watchdog)" | grep -v grep
或配置 launchd 开机自启（macOS）：
cat > ~/Library/LaunchAgents/com.zz-chat.bridge.plist << PLIST
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
<key>Label</key><string>com.zz-chat.bridge</string>
<key>ProgramArguments</key>
<array>
<string>/usr/local/bin/python3</string><string>-u</string>
<string>/Users/$(whoami)/.openclaw/workspace/openchat/bridge/watchdog.py</string>
</array>
<key>EnvironmentVariables</key>
<dict>
<key>PATH</key><string>/Users/$(whoami)/.local/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin</string>
<key>all_proxy</key><string></string><key>ALL_PROXY</key><string></string>
<key>http_proxy</key><string></string><key>HTTP_PROXY</key><string></string>
<key>https_proxy</key><string></string><key>HTTPS_PROXY</key><string></string>
</dict>
<key>RunAtLoad</key><true/><key>KeepAlive</key><true/>
<key>StandardOutPath</key><string>/Users/$(whoami)/.openclaw/workspace/openchat/bridge/watchdog.log</string>
<key>StandardErrorPath</key><string>/Users/$(whoami)/.openclaw/workspace/openchat/bridge/watchdog.log</string>
<key>WorkingDirectory</key><string>/Users/$(whoami)/.openclaw/workspace/openchat/bridge</string>
</dict>
</plist>
PLIST
launchctl load ~/Library/LaunchAgents/com.zz-chat.bridge.plist
sleep 5
ps aux | grep -E "(bridge|watchdog)" | grep -v grep
用 canvas 截图二维码发送给用户：
ZZ_ID=$(cat ~/.zz/id)
WORKER_URL=$(cat ~/.zz/worker_url)
PAGE_URL=https://badxtdss.github.io/zz-chat/?bridge=${ZZ_ID}&api=$(python3 -c "import urllib.parse; print(urllib.parse.quote('$WORKER_URL', safe=''))")
保存 HTML 到 /tmp/zz-qr.html：
<!DOCTYPE html>
<html>
<head><meta charset="UTF-8"><title>爪爪</title>
<style>
body{font-family:-apple-system,sans-serif;background:#0a0a0a;color:#e0e0e0;display:flex;flex-direction:column;align-items:center;justify-content:center;height:100vh;margin:0}
h1{font-size:24px;margin-bottom:8px}
.id{font-size:48px;font-weight:bold;color:#4fc3f7;margin:12px 0}
.qr{background:#fff;padding:12px;border-radius:12px;margin:16px 0}
.hint{color:#8892b0;:;:center;:;:}
🦞 爪爪
编号 {ZZ_ID}
手机扫码打开爪爪消息通过你的 OpenClaw 处理
用 canvas 工具截图（action="present" → action="snapshot"），然后发图 + 文字消息：
✅ 爪爪已部署完成！
🦞 你的编号: {ZZ_ID}
🔗 链接: {PAGE_URL}
📱 添加到桌面（必做）:
1. 打开后点浏览器菜单 → 「访问电脑版」
2. 再点菜单 → 「添加到桌面」
微信/QQ: 右上角 ··· → 「添加到桌面」
Safari: 分享按钮 → 「添加到主屏幕」
3. 以后从桌面图标打开，体验最佳
| 聊天类型 | 路径 | Worker |
|---|---|---|
| OC（OpenClaw） | 手机→Worker→bridge→CLI | 用户自己的 Worker |
| 朋友互聊（在线） | 手机A→WorkerA→WorkerB→手机B | 两个用户的 Worker |
| 朋友互聊（离线文字） | WorkerB 存 DO 持久存储，上线补发 | 接收方的 Worker |
| 注册 | Worker→中心Worker（ai0000.cn） | 中心 Worker |
| 文件 | 用途 |
|---|---|
worker.js | 中心 Worker（ai0000.cn 用） |
worker-user.js | 用户独立 Worker |
wrangler.toml | 中心 Worker 配置 |
wrangler-user.toml | 用户 Worker 配置模板 |
bridge.py | 桥接脚本，支持 --worker 和 --uid |
watchdog.py | 看门狗，监控 bridge |
index.html | 手机端首页 |
chat.html | 手机端 OC 聊天页（WebSocket） |
bridge.js | Node.js 版桥接（Windows 兼容） |
start-bridge.bat | Windows 启动脚本 |
bridge.py --worker <URL> # 指定 Worker 地址
bridge.py --uid <ID> # 直接指定编号
launchd → watchdog.py → bridge.py
（系统级） （进程级） （实际桥接）
?bridge= + ?api= 参数）才能进入网页openclaw agent CLI 处理~/.openclaw/workspace/openchat/bridge/watchdog.log~/.openclaw/workspace/openchat/bridge/bridge.log🦞 爪爪 by 秋风悠扬