1bf5402d34
- 3-create-persistence.ps1: chain dd && mkfs.ext4 with && (was ;, so dd
failures still ran mkfs and produced bogus images); fix \wsl$\$wslDistro
variable expansion ambiguity using \${wslDistro}; mirror docker-desktop
branch with the same && chaining
- setup-openclaw.sh: guard mktemp failure; ensure partial node tarballs are
cleaned via EXIT trap; tolerate hosts with no non-loopback NIC under
pipefail; don't let test-installation.sh failure exit the post-install
- install.sh: add `set -o pipefail`; surface npm install failures (was
hidden by `| tail -5`); JSON-escape user-provided API_KEY before writing
to openclaw.json so keys with backslashes/quotes don't break the config
- Add SECURITY.md: private disclosure channel + scope
1168 lines
35 KiB
Bash
Executable File
1168 lines
35 KiB
Bash
Executable File
#!/bin/bash
|
||
# ============================================================
|
||
# U-Claw 一键安装脚本 (Mac/Linux)
|
||
# 用法: curl -fsSL https://u-claw.org/install.sh | bash
|
||
# 或: bash install.sh
|
||
# ============================================================
|
||
|
||
set -e
|
||
set -o pipefail
|
||
# 注意：未启用 `set -u` —— 第 5 步交互式分支会有意把 API_KEY/BASE_URL/KEY_LABEL 等未赋值的变量
|
||
# 用 [-z "$X"] 检测后再写配置；启用 -u 会破坏这条降级路径。改用显式 `: "${VAR:=}"` 默认值过于侵入。
|
||
|
||
# ---- 颜色定义 ----
|
||
GREEN='\033[0;32m'
|
||
CYAN='\033[0;36m'
|
||
RED='\033[0;31m'
|
||
YELLOW='\033[1;33m'
|
||
BOLD='\033[1m'
|
||
DIM='\033[2m'
|
||
NC='\033[0m'
|
||
|
||
# ---- 常量 ----
|
||
UCLAW_DIR="$HOME/.uclaw"
|
||
RUNTIME_DIR="$UCLAW_DIR/runtime"
|
||
CORE_DIR="$UCLAW_DIR/core"
|
||
DATA_DIR="$UCLAW_DIR/data"
|
||
CONFIG_PATH="$DATA_DIR/.openclaw/openclaw.json"
|
||
NODE_VERSION="v22.16.0"
|
||
MIRROR="https://registry.npmmirror.com"
|
||
NODE_MIRROR="https://npmmirror.com/mirrors/node"
|
||
|
||
# ============================================================
|
||
# Step 1: Banner + 系统检测
|
||
# ============================================================
|
||
clear 2>/dev/null || true
|
||
echo ""
|
||
echo -e "${CYAN}${BOLD}"
|
||
cat << 'BANNER'
|
||
╔══════════════════════════════════════════╗
|
||
║ 🦞 U-Claw 一键安装 ║
|
||
║ 让 AI 助手一行命令装好 ║
|
||
╚══════════════════════════════════════════╝
|
||
BANNER
|
||
echo -e "${NC}"
|
||
|
||
# 系统检测
|
||
OS=$(uname -s)
|
||
ARCH=$(uname -m)
|
||
|
||
if [ "$OS" = "Darwin" ]; then
|
||
if [ "$ARCH" = "arm64" ]; then
|
||
PLATFORM="darwin-arm64"
|
||
NODE_DIR="node-mac-arm64"
|
||
echo -e " 系统: ${GREEN}macOS Apple Silicon (M 系列) ✓${NC}"
|
||
else
|
||
PLATFORM="darwin-x64"
|
||
NODE_DIR="node-mac-x64"
|
||
echo -e " 系统: ${GREEN}macOS Intel ✓${NC}"
|
||
fi
|
||
elif [ "$OS" = "Linux" ]; then
|
||
if [ "$ARCH" = "x86_64" ]; then
|
||
PLATFORM="linux-x64"
|
||
NODE_DIR="node-linux-x64"
|
||
echo -e " 系统: ${GREEN}Linux x64 ✓${NC}"
|
||
elif [ "$ARCH" = "aarch64" ] || [ "$ARCH" = "arm64" ]; then
|
||
PLATFORM="linux-arm64"
|
||
NODE_DIR="node-linux-arm64"
|
||
echo -e " 系统: ${GREEN}Linux ARM64 ✓${NC}"
|
||
else
|
||
echo -e " ${RED}不支持的架构: $ARCH${NC}"
|
||
exit 1
|
||
fi
|
||
else
|
||
echo -e " ${RED}不支持的系统: $OS${NC}"
|
||
echo -e " ${YELLOW}Windows 请使用 PowerShell: irm https://u-claw.org/install.ps1 | iex${NC}"
|
||
exit 1
|
||
fi
|
||
|
||
echo -e " 安装目录: ${CYAN}$UCLAW_DIR${NC}"
|
||
echo ""
|
||
|
||
# 检查已有安装
|
||
if [ -d "$UCLAW_DIR/core/node_modules/openclaw" ]; then
|
||
echo -e " ${YELLOW}检测到已有安装: $UCLAW_DIR${NC}"
|
||
if [ -t 0 ]; then
|
||
read -p " 覆盖安装？(y/n) [y]: " -n 1 OVERWRITE
|
||
echo ""
|
||
if [ "$OVERWRITE" = "n" ] || [ "$OVERWRITE" = "N" ]; then
|
||
echo -e " ${DIM}已取消${NC}"
|
||
exit 0
|
||
fi
|
||
else
|
||
echo -e " ${CYAN}管道模式，自动覆盖安装${NC}"
|
||
fi
|
||
echo ""
|
||
fi
|
||
|
||
# 创建目录结构
|
||
mkdir -p "$RUNTIME_DIR" "$CORE_DIR" "$DATA_DIR/.openclaw" "$DATA_DIR/memory" "$DATA_DIR/backups"
|
||
|
||
# ============================================================
|
||
# Step 2: Node.js v22 安装
|
||
# ============================================================
|
||
echo -e " ${BOLD}[1/7] 安装 Node.js $NODE_VERSION ...${NC}"
|
||
|
||
NODE_INSTALL_DIR="$RUNTIME_DIR/$NODE_DIR"
|
||
INSTALL_NODE=""
|
||
INSTALL_NPM=""
|
||
|
||
# 检查系统 Node.js
|
||
USE_SYSTEM_NODE=false
|
||
if command -v node >/dev/null 2>&1; then
|
||
SYS_VER=$(node --version)
|
||
MAJOR=$(echo "$SYS_VER" | sed 's/v//' | cut -d. -f1)
|
||
if [ "$MAJOR" -ge 20 ] 2>/dev/null; then
|
||
echo -e " ${GREEN}✓${NC} 系统已有 Node.js $SYS_VER，复用"
|
||
INSTALL_NODE="$(which node)"
|
||
INSTALL_NPM="$(which npm)"
|
||
USE_SYSTEM_NODE=true
|
||
fi
|
||
fi
|
||
|
||
if [ "$USE_SYSTEM_NODE" = "false" ]; then
|
||
if [ -f "$NODE_INSTALL_DIR/bin/node" ]; then
|
||
echo -e " ${GREEN}✓${NC} Node.js 已存在，跳过下载"
|
||
INSTALL_NODE="$NODE_INSTALL_DIR/bin/node"
|
||
INSTALL_NPM="$NODE_INSTALL_DIR/bin/npm-cli.js"
|
||
else
|
||
echo -e " ${CYAN}↓${NC} 从国内镜像下载 Node.js $NODE_VERSION ($PLATFORM)..."
|
||
TARBALL="node-${NODE_VERSION}-${PLATFORM}.tar.gz"
|
||
URL="${NODE_MIRROR}/${NODE_VERSION}/${TARBALL}"
|
||
|
||
mkdir -p "$NODE_INSTALL_DIR"
|
||
if command -v curl >/dev/null 2>&1; then
|
||
curl -# -L "$URL" -o "/tmp/$TARBALL"
|
||
elif command -v wget >/dev/null 2>&1; then
|
||
wget -q --show-progress "$URL" -O "/tmp/$TARBALL"
|
||
else
|
||
echo -e " ${RED}✗ 未找到 curl 或 wget，请先安装${NC}"
|
||
exit 1
|
||
fi
|
||
|
||
tar -xzf "/tmp/$TARBALL" -C "$NODE_INSTALL_DIR" --strip-components=1
|
||
rm -f "/tmp/$TARBALL"
|
||
chmod +x "$NODE_INSTALL_DIR/bin/node"
|
||
|
||
if [ -f "$NODE_INSTALL_DIR/bin/node" ]; then
|
||
echo -e " ${GREEN}✓${NC} Node.js 安装完成"
|
||
INSTALL_NODE="$NODE_INSTALL_DIR/bin/node"
|
||
INSTALL_NPM="$NODE_INSTALL_DIR/lib/node_modules/npm/bin/npm-cli.js"
|
||
else
|
||
echo -e " ${RED}✗ Node.js 下载失败${NC}"
|
||
exit 1
|
||
fi
|
||
fi
|
||
fi
|
||
|
||
# Helper: invoke npm. For bundled Node we have npm-cli.js (run via node).
|
||
# For system Node, INSTALL_NPM is the npm shell wrapper (executable).
|
||
run_npm() {
|
||
if [ "$USE_SYSTEM_NODE" = "true" ]; then
|
||
"$INSTALL_NPM" "$@"
|
||
else
|
||
"$INSTALL_NODE" "$INSTALL_NPM" "$@"
|
||
fi
|
||
}
|
||
|
||
echo ""
|
||
|
||
# ============================================================
|
||
# Step 3: OpenClaw 安装
|
||
# ============================================================
|
||
echo -e " ${BOLD}[2/7] 安装 OpenClaw ...${NC}"
|
||
|
||
if [ -d "$CORE_DIR/node_modules/openclaw" ]; then
|
||
echo -e " ${GREEN}✓${NC} OpenClaw 已安装，跳过"
|
||
else
|
||
if [ ! -f "$CORE_DIR/package.json" ]; then
|
||
cat > "$CORE_DIR/package.json" << 'PKGJSON'
|
||
{
|
||
"name": "u-claw-core",
|
||
"version": "1.0.0",
|
||
"private": true,
|
||
"dependencies": {
|
||
"openclaw": "latest"
|
||
}
|
||
}
|
||
PKGJSON
|
||
fi
|
||
|
||
echo -e " ${CYAN}↓${NC} 从国内镜像安装..."
|
||
NPM_LOG=$(mktemp /tmp/uclaw-npm.XXXXXX.log)
|
||
if run_npm install --prefix "$CORE_DIR" --registry="$MIRROR" >"$NPM_LOG" 2>&1; then
|
||
tail -5 "$NPM_LOG"
|
||
rm -f "$NPM_LOG"
|
||
echo -e " ${GREEN}✓${NC} OpenClaw 安装完成"
|
||
else
|
||
echo -e " ${RED}✗ OpenClaw 安装失败,完整日志见 $NPM_LOG${NC}"
|
||
tail -20 "$NPM_LOG"
|
||
exit 1
|
||
fi
|
||
fi
|
||
|
||
echo ""
|
||
|
||
# ============================================================
|
||
# Step 4: QQ 插件（非致命）
|
||
# ============================================================
|
||
echo -e " ${BOLD}[3/7] 安装 QQ 插件 ...${NC}"
|
||
|
||
if [ -d "$CORE_DIR/node_modules/@sliverp/qqbot" ]; then
|
||
echo -e " ${GREEN}✓${NC} QQ 插件已安装，跳过"
|
||
else
|
||
echo -e " ${CYAN}↓${NC} 安装 QQ 插件..."
|
||
run_npm install @sliverp/qqbot@latest --prefix "$CORE_DIR" --registry="$MIRROR" 2>/dev/null || {
|
||
echo -e " ${YELLOW}⚠${NC} QQ 插件安装失败（不影响主功能）"
|
||
}
|
||
if [ -d "$CORE_DIR/node_modules/@sliverp/qqbot" ]; then
|
||
echo -e " ${GREEN}✓${NC} QQ 插件安装完成"
|
||
fi
|
||
fi
|
||
|
||
echo ""
|
||
|
||
# ============================================================
|
||
# Step 5: 写入 10 个中国技能
|
||
# ============================================================
|
||
echo -e " ${BOLD}[4/7] 安装中国本地化技能 (10个) ...${NC}"
|
||
|
||
SKILLS_TARGET="$CORE_DIR/node_modules/openclaw/skills"
|
||
if [ ! -d "$SKILLS_TARGET" ]; then
|
||
mkdir -p "$SKILLS_TARGET"
|
||
fi
|
||
|
||
SKILL_COUNT=0
|
||
|
||
# ---- bilibili-helper ----
|
||
if [ ! -d "$SKILLS_TARGET/bilibili-helper" ]; then
|
||
mkdir -p "$SKILLS_TARGET/bilibili-helper"
|
||
cat > "$SKILLS_TARGET/bilibili-helper/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: bilibili-helper
|
||
description: "B站内容助手 - 视频标题描述优化、标签策略、封面设计建议、分区选择、评论互动"
|
||
metadata: { "openclaw": { "emoji": "📺" } }
|
||
---
|
||
|
||
# B站内容助手
|
||
|
||
帮助 UP 主优化视频标题、描述、标签和封面，提升视频在 B 站的推荐和互动表现。
|
||
|
||
## 功能概述
|
||
|
||
- **标题优化**: 符合 B 站用户口味的标题写法
|
||
- **描述撰写**: 含时间轴、关键词、引导语的视频简介
|
||
- **标签策略**: 精准标签 + 热门标签组合，提升推荐
|
||
- **分区建议**: 根据内容推荐最佳投稿分区
|
||
- **封面设计**: 封面构图、文字、配色建议
|
||
- **评论互动**: 置顶评论话术，提升互动率
|
||
|
||
## 使用示例
|
||
|
||
### 写标题和描述
|
||
```
|
||
我做了一个"用AI写代码"的教程视频，帮我写标题和描述
|
||
```
|
||
|
||
### 分区推荐
|
||
```
|
||
我的视频内容是游戏+编程结合，应该投哪个分区？
|
||
```
|
||
|
||
### 封面建议
|
||
```
|
||
视频主题是"大学生省钱攻略"，帮我设计封面方案
|
||
```
|
||
|
||
## 标题公式
|
||
|
||
1. **疑问式**: "为什么XX？看完你就懂了"
|
||
2. **教程式**: "XX教程｜从零开始手把手教学"
|
||
3. **测评式**: "花了XX元买了XX，值不值？"
|
||
4. **挑战式**: "挑战XX天只用XX"
|
||
5. **盘点式**: "XX年度十大XX盘点"
|
||
6. **震惊式**: "这也太XX了吧！"（B站经典体）
|
||
|
||
## 描述模板
|
||
|
||
```
|
||
【视频简介】
|
||
一句话概括视频内容
|
||
|
||
【时间轴】
|
||
00:00 开场
|
||
01:23 第一部分
|
||
05:00 第二部分
|
||
08:30 总结
|
||
|
||
【相关链接】
|
||
工具/资源链接
|
||
|
||
【关于我】
|
||
一句话介绍 + 更新频率
|
||
|
||
三连支持一下吧~ 有问题评论区见！
|
||
```
|
||
|
||
## 分区选择指南
|
||
|
||
| 内容类型 | 推荐分区 |
|
||
|---------|---------|
|
||
| 编程教程 | 科技 → 计算机技术 |
|
||
| 日常 vlog | 生活 → 日常 |
|
||
| 游戏实况 | 游戏 → 单机/网游 |
|
||
| 知识科普 | 知识 → 社科人文/科学科普 |
|
||
| 美食制作 | 生活 → 美食圈 |
|
||
| AI/数码 | 科技 → 软件应用 |
|
||
|
||
## 标签策略
|
||
|
||
- **核心标签**: 精确描述视频内容（2-3个）
|
||
- **分区热门**: 当前分区下的热门标签（2-3个）
|
||
- **全站热门**: B站当前热点话题（1-2个）
|
||
- **长尾标签**: 细分领域关键词（2-3个）
|
||
- 总计 8-12 个标签效果最佳
|
||
|
||
## 封面设计要点
|
||
|
||
- **尺寸**: 16:10（推荐 1146x717）
|
||
- **文字**: 不超过 10 字，字体加粗加描边
|
||
- **表情**: 人物表情夸张更吸引点击
|
||
- **配色**: 高饱和度，与B站粉色/蓝色背景形成对比
|
||
- **避免**: 过多文字、低分辨率、与内容无关的封面
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- china-search ----
|
||
if [ ! -d "$SKILLS_TARGET/china-search" ]; then
|
||
mkdir -p "$SKILLS_TARGET/china-search"
|
||
cat > "$SKILLS_TARGET/china-search/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: china-search
|
||
description: "国内搜索引擎 - 百度、搜狗、Bing中国搜索，绕过GFW限制"
|
||
metadata: { "openclaw": { "emoji": "🔍" } }
|
||
---
|
||
|
||
# 国内搜索引擎助手
|
||
|
||
在无法访问 Google 的环境下，通过 curl 调用百度、搜狗、Bing 中国等国内搜索引擎获取信息。
|
||
|
||
## 功能概述
|
||
|
||
- **百度搜索**: 中文信息最全的搜索引擎
|
||
- **搜狗搜索**: 微信公众号文章搜索利器
|
||
- **Bing 中国**: 国际信息 + 国内可访问
|
||
|
||
## 搜索命令
|
||
|
||
### 百度搜索
|
||
```bash
|
||
curl -s -L "https://www.baidu.com/s?wd=关键词" \
|
||
-H "User-Agent: Mozilla/5.0" | head -200
|
||
```
|
||
|
||
### 搜狗搜索（含微信文章）
|
||
```bash
|
||
curl -s -L "https://weixin.sogou.com/weixin?query=关键词" \
|
||
-H "User-Agent: Mozilla/5.0"
|
||
```
|
||
|
||
### Bing 中国搜索
|
||
```bash
|
||
curl -s -L "https://cn.bing.com/search?q=关键词" \
|
||
-H "User-Agent: Mozilla/5.0"
|
||
```
|
||
|
||
## 搜索技巧
|
||
|
||
| 语法 | 说明 | 示例 |
|
||
|-----|------|------|
|
||
| "" | 精确匹配 | "人工智能" |
|
||
| site: | 限定网站 | site:zhihu.com AI |
|
||
| filetype: | 限定文件类型 | filetype:pdf 机器学习 |
|
||
| - | 排除关键词 | 苹果 -手机 |
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- china-translate ----
|
||
if [ ! -d "$SKILLS_TARGET/china-translate" ]; then
|
||
mkdir -p "$SKILLS_TARGET/china-translate"
|
||
cat > "$SKILLS_TARGET/china-translate/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: china-translate
|
||
description: "中英互译 + 本地化 - 技术翻译、UI本地化、文化适配、避免机翻腔"
|
||
metadata: { "openclaw": { "emoji": "🌐" } }
|
||
---
|
||
|
||
# 中英互译 + 本地化助手
|
||
|
||
专业的中英双向翻译工具，专注技术文档翻译、产品本地化和文化适配，杜绝机翻腔。
|
||
|
||
## 功能概述
|
||
|
||
- **技术翻译**: 编程/AI/互联网领域的专业术语翻译
|
||
- **UI 本地化**: 将英文界面文案翻译为符合中文习惯的表达
|
||
- **文化适配**: 处理中西文化差异，避免直译造成的误解
|
||
- **去机翻腔**: 将僵硬的机器翻译改写为自然流畅的中文
|
||
|
||
## 常见术语对照表
|
||
|
||
| English | 中文（推荐） | 避免使用 |
|
||
|---------|------------|---------|
|
||
| Machine Learning | 机器学习 | 机械学习 |
|
||
| Deploy | 部署 | 配置/展开 |
|
||
| Repository | 仓库 / 代码仓库 | 存储库 |
|
||
| Pull Request | PR / 合并请求 | 拉取请求 |
|
||
| Container | 容器 | 集装箱 |
|
||
| Middleware | 中间件 | 中间软件 |
|
||
| Render | 渲染 | 呈现 |
|
||
| Token | Token / 令牌 | 代币（非区块链语境） |
|
||
| Prompt | 提示词 / Prompt | 提示语 |
|
||
| Fine-tune | 微调 | 精调 |
|
||
|
||
## UI 本地化指南
|
||
|
||
- "Submit" → "提交"
|
||
- "Cancel" → "取消"
|
||
- "Get Started" → "立即开始"
|
||
- "Learn More" → "了解更多"
|
||
- "Sign Up" → "注册"
|
||
- 中文 UI 不需要句号（按钮、标签）
|
||
- 中英文之间加半角空格
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- china-weather ----
|
||
if [ ! -d "$SKILLS_TARGET/china-weather" ]; then
|
||
mkdir -p "$SKILLS_TARGET/china-weather"
|
||
cat > "$SKILLS_TARGET/china-weather/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: china-weather
|
||
description: "中国城市天气查询 - 支持中文城市名、农历日期、wttr.in接口"
|
||
metadata: { "openclaw": { "emoji": "🌤️" } }
|
||
---
|
||
|
||
# 中国城市天气查询
|
||
|
||
通过 wttr.in 查询中国各城市天气信息，支持中文城市名输入。
|
||
|
||
## 查询方式
|
||
|
||
```bash
|
||
# 基础查询（中文输出）
|
||
curl -s "wttr.in/深圳?lang=zh"
|
||
|
||
# 精简单行格式
|
||
curl -s "wttr.in/深圳?format=3&lang=zh"
|
||
|
||
# JSON 格式（便于解析）
|
||
curl -s "wttr.in/深圳?format=j1"
|
||
```
|
||
|
||
## 使用示例
|
||
|
||
- 深圳今天天气怎么样？
|
||
- 明天要去杭州出差，需要带伞吗？
|
||
- 北京和上海这周末天气哪个好？
|
||
|
||
## 注意事项
|
||
|
||
- wttr.in 为免费服务，偶尔可能响应较慢
|
||
- 温度默认摄氏度，风速默认 km/h
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- deepseek-helper ----
|
||
if [ ! -d "$SKILLS_TARGET/deepseek-helper" ]; then
|
||
mkdir -p "$SKILLS_TARGET/deepseek-helper"
|
||
cat > "$SKILLS_TARGET/deepseek-helper/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: deepseek-helper
|
||
description: "DeepSeek API 助手 - 编程辅助、模型选择、API调用指南、定价信息"
|
||
metadata: { "openclaw": { "emoji": "🤖" } }
|
||
---
|
||
|
||
# DeepSeek API 助手
|
||
|
||
帮助用户高效使用 DeepSeek 系列模型，包括模型选择建议、API 调用示例和定价计算。
|
||
|
||
## 模型对比
|
||
|
||
| 模型 | 适用场景 | 上下文长度 |
|
||
|------|---------|-----------|
|
||
| deepseek-chat | 日常对话、文案写作、知识问答 | 32K |
|
||
| deepseek-coder | 代码生成、代码审查、技术文档 | 16K |
|
||
| deepseek-reasoner | 复杂推理、数学、逻辑分析 | 64K |
|
||
|
||
## 关键提示
|
||
|
||
- DeepSeek API 兼容 OpenAI 格式，base_url 改为 `https://api.deepseek.com`
|
||
- 国内访问无需翻墙，延迟低
|
||
- 支持 function calling 和 JSON mode
|
||
- API Key 申请地址: https://platform.deepseek.com
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- douyin-script ----
|
||
if [ ! -d "$SKILLS_TARGET/douyin-script" ]; then
|
||
mkdir -p "$SKILLS_TARGET/douyin-script"
|
||
cat > "$SKILLS_TARGET/douyin-script/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: douyin-script
|
||
description: "抖音/快手短视频脚本 - 前3秒hook、脚本结构、热门音乐建议、话题标签策略"
|
||
metadata: { "openclaw": { "emoji": "🎬" } }
|
||
---
|
||
|
||
# 抖音/快手短视频脚本助手
|
||
|
||
帮你写出完播率高、互动强的短视频脚本。
|
||
|
||
## 脚本结构模板
|
||
|
||
### 15秒脚本
|
||
```
|
||
【0-3秒】Hook: 抛出反常识/痛点
|
||
【3-10秒】核心内容: 快速讲解
|
||
【10-15秒】CTA: 引导互动
|
||
```
|
||
|
||
### 60秒脚本
|
||
```
|
||
【0-3秒】Hook: 强烈的情绪/悬念
|
||
【3-15秒】背景铺垫
|
||
【15-40秒】内容展开
|
||
【40-50秒】反转/高潮
|
||
【50-60秒】CTA + 话题标签
|
||
```
|
||
|
||
## 前3秒 Hook 公式
|
||
|
||
1. **反常识**: "你一直在做的XX其实是错的"
|
||
2. **数字冲击**: "只花了100块，效果比1000块的还好"
|
||
3. **悬念提问**: "猜猜这个东西是干什么的？"
|
||
4. **情绪共鸣**: "打工人看完都沉默了..."
|
||
5. **结果前置**: "最终效果太绝了！往下看做法"
|
||
|
||
## 话题标签策略
|
||
|
||
- **大话题**: #抖音热门 #今日份分享（1-2个）
|
||
- **中话题**: #健身打卡 #美食教程（2-3个）
|
||
- **小话题**: #深圳探店 #程序员日常（2-3个）
|
||
|
||
## 发布建议
|
||
|
||
- **黄金时段**: 早7-9点、午12-14点、晚18-22点
|
||
- **封面**: 竖屏 9:16，文字大且少（不超过8字）
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- wechat-article ----
|
||
if [ ! -d "$SKILLS_TARGET/wechat-article" ]; then
|
||
mkdir -p "$SKILLS_TARGET/wechat-article"
|
||
cat > "$SKILLS_TARGET/wechat-article/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: wechat-article
|
||
description: "微信公众号文章写作 - 文章结构、排版规范、阅读转化优化、封面建议"
|
||
metadata: { "openclaw": { "emoji": "💚" } }
|
||
---
|
||
|
||
# 微信公众号文章写作助手
|
||
|
||
帮助运营者写出高阅读、高转发的微信公众号文章。
|
||
|
||
## 文章结构模板
|
||
|
||
```
|
||
【标题】控制在 30 字以内，前 15 字抓眼球
|
||
【摘要】出现在分享卡片，控制在 54 字
|
||
|
||
【引言】1-2 段，建立共鸣或抛出问题
|
||
|
||
【正文】
|
||
## 小标题一
|
||
内容段落...（每段 3-5 行）
|
||
|
||
## 小标题二
|
||
内容段落...
|
||
|
||
【总结】回扣主题，升华观点
|
||
|
||
【尾部】
|
||
觉得有收获？点个「在看」让更多人看到
|
||
```
|
||
|
||
## 微信排版规范
|
||
|
||
- **正文字号**: 15-16px
|
||
- **行间距**: 1.75-2倍
|
||
- **段间距**: 空一行
|
||
- **配色**: 正文 #3f3f3f，强调 #007AFF
|
||
- **图片**: 宽度 100%，JPG，单张不超过 5MB
|
||
|
||
## 标题技巧
|
||
|
||
1. **数字开头**: "3个方法/5分钟学会"
|
||
2. **疑问式**: "为什么XX？答案出乎意料"
|
||
3. **对比式**: "XX和XX的差距，在于这一点"
|
||
4. **紧迫感**: "再不XX就晚了"
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- weibo-poster ----
|
||
if [ ! -d "$SKILLS_TARGET/weibo-poster" ]; then
|
||
mkdir -p "$SKILLS_TARGET/weibo-poster"
|
||
cat > "$SKILLS_TARGET/weibo-poster/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: weibo-poster
|
||
description: "微博内容创作 - 140字优化、话题热搜、@提及策略、配图描述、发布时机"
|
||
metadata: { "openclaw": { "emoji": "🔴" } }
|
||
---
|
||
|
||
# 微博内容创作助手
|
||
|
||
帮你写出高转发、高互动的微博内容。
|
||
|
||
## 微博内容模板
|
||
|
||
### 短微博（140字以内）
|
||
```
|
||
【观点/金句】一句话表达核心观点
|
||
【补充说明】1-2句展开
|
||
【话题标签】#话题一# #话题二#
|
||
```
|
||
|
||
## 140字写作技巧
|
||
|
||
- **先写后删**: 先完整表达，再精简到 140 字
|
||
- **一条一观点**: 不要在一条微博里讲太多
|
||
- **金句收尾**: 最后一句要值得被转发
|
||
|
||
## 配图建议
|
||
|
||
- **1张图**: 重点突出，适合海报/截图
|
||
- **3张图**: 对比/过程展示
|
||
- **6张图**: 故事叙述
|
||
- **9张图**: 九宫格，视觉冲击力最强
|
||
|
||
## 发布时间建议
|
||
|
||
- **工作日**: 午休 12:00-13:00、下班后 18:00-20:00
|
||
- **周末**: 上午 10:00-11:00、晚上 20:00-22:00
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- xiaohongshu-writer ----
|
||
if [ ! -d "$SKILLS_TARGET/xiaohongshu-writer" ]; then
|
||
mkdir -p "$SKILLS_TARGET/xiaohongshu-writer"
|
||
cat > "$SKILLS_TARGET/xiaohongshu-writer/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: xiaohongshu-writer
|
||
description: "小红书笔记写作助手 - 标题优化、emoji策略、话题标签、封面文案、笔记结构"
|
||
metadata: { "openclaw": { "emoji": "📕" } }
|
||
---
|
||
|
||
# 小红书笔记写作助手
|
||
|
||
专业的小红书内容创作工具，帮你写出高互动、高收藏的爆款笔记。
|
||
|
||
## 笔记结构模板
|
||
|
||
```
|
||
【标题】数字 + 痛点/好奇心 + emoji
|
||
例: 打工人必看！5个让老板刮目相看的汇报技巧 💼
|
||
|
||
【开头 Hook】1-2句直击痛点
|
||
【正文】分点罗列，每点 2-3 行
|
||
1️⃣ 第一点...
|
||
2️⃣ 第二点...
|
||
|
||
【结尾 CTA】
|
||
觉得有用就 ❤️ 收藏起来慢慢看！
|
||
|
||
【话题标签】
|
||
#职场干货 #汇报技巧 #打工人必看
|
||
```
|
||
|
||
## 标题公式
|
||
|
||
1. **数字法**: "5个/10种/100元以内" + 关键词
|
||
2. **反差法**: "月薪3千 vs 月薪3万的区别"
|
||
3. **身份法**: "作为XX人/XX年经验告诉你"
|
||
4. **测评法**: "亲测有效！" + 结果
|
||
5. **合集法**: "XX合集｜一篇搞定"
|
||
|
||
## 写作要点
|
||
|
||
- 每段不超过 3 行
|
||
- emoji 不要堆砌，每段 1-2 个
|
||
- 话题标签 3-8 个为佳
|
||
- 封面文案控制在 10 字以内
|
||
- 发布时间: 早7-9点、午12-14点、晚20-22点
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
# ---- zhihu-writer ----
|
||
if [ ! -d "$SKILLS_TARGET/zhihu-writer" ]; then
|
||
mkdir -p "$SKILLS_TARGET/zhihu-writer"
|
||
cat > "$SKILLS_TARGET/zhihu-writer/SKILL.md" << 'SKILLEOF'
|
||
---
|
||
name: zhihu-writer
|
||
description: "知乎回答/文章写作 - 回答结构、专业语气、引用规范、话题关联、盐值优化"
|
||
metadata: { "openclaw": { "emoji": "📝" } }
|
||
---
|
||
|
||
# 知乎回答/文章写作助手
|
||
|
||
帮助创作者写出高赞、高收藏的知乎内容。
|
||
|
||
## 回答结构模板
|
||
|
||
```
|
||
【开头】直接回答问题 + 亮明立场
|
||
"作为一个做了5年XX的人，我的回答是：..."
|
||
|
||
---
|
||
|
||
【正文】分点论述，有理有据
|
||
一、第一个论点 + 数据/案例
|
||
二、第二个论点 + 个人经验
|
||
三、第三个论点 + 行业分析
|
||
|
||
【总结】回扣问题，给出实操建议
|
||
```
|
||
|
||
## 知乎写作风格
|
||
|
||
- **自信但不傲慢**: "以我的经验来看"
|
||
- **专业但易懂**: 术语需要解释
|
||
- **有态度但包容**: 尊重不同立场
|
||
- **真诚分享**: 多用个人经验和案例
|
||
|
||
## 盐值优化
|
||
|
||
- 回答字数建议 500-3000 字
|
||
- 避免复制粘贴，用自己的语言
|
||
- 回复评论，参与讨论
|
||
- 持续在 2-3 个话题下创作
|
||
SKILLEOF
|
||
SKILL_COUNT=$((SKILL_COUNT + 1))
|
||
fi
|
||
|
||
echo -e " ${GREEN}✓${NC} 中国技能安装完成 (+${SKILL_COUNT} 个)"
|
||
echo ""
|
||
|
||
# ============================================================
|
||
# Step 6: 交互式模型配置
|
||
# ============================================================
|
||
echo -e " ${BOLD}[5/7] 配置 AI 模型 ...${NC}"
|
||
echo ""
|
||
|
||
# 如果已有配置且包含 apiKey，跳过
|
||
if [ -f "$CONFIG_PATH" ] && grep -q '"apiKey"' "$CONFIG_PATH" 2>/dev/null; then
|
||
echo -e " ${GREEN}✓${NC} 已有模型配置，跳过"
|
||
else
|
||
# 检测是否可交互
|
||
if [ -t 0 ]; then
|
||
echo -e " 请选择 AI 模型:"
|
||
echo ""
|
||
echo -e " ${BOLD}── 国内推荐（无需翻墙）──${NC}"
|
||
echo -e " ${GREEN}1)${NC} DeepSeek ⭐ 推荐，性价比最高"
|
||
echo -e " ${NC}2)${NC} Kimi/月之暗面"
|
||
echo -e " ${NC}3)${NC} 通义千问/阿里"
|
||
echo -e " ${NC}4)${NC} 智谱GLM"
|
||
echo -e " ${NC}5)${NC} MiniMax"
|
||
echo -e " ${NC}6)${NC} 豆包/火山引擎"
|
||
echo -e " ${NC}7)${NC} 硅基流动"
|
||
echo ""
|
||
echo -e " ${BOLD}── 海外模型 ──${NC}"
|
||
echo -e " ${NC}8)${NC} Claude ${NC}9)${NC} GPT"
|
||
echo ""
|
||
echo -e " ${BOLD}── 本地 ──${NC}"
|
||
echo -e " ${NC}10)${NC} Ollama 本地模型"
|
||
echo ""
|
||
read -p " 请输入编号 [1]: " MODEL_CHOICE
|
||
MODEL_CHOICE=${MODEL_CHOICE:-1}
|
||
|
||
# 模型配置映射
|
||
case $MODEL_CHOICE in
|
||
1)
|
||
MODEL_NAME="deepseek-chat"
|
||
BASE_URL="https://api.deepseek.com/v1"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="DeepSeek API Key"
|
||
KEY_HINT="获取地址: https://platform.deepseek.com/api_keys"
|
||
NEED_KEY=true
|
||
;;
|
||
2)
|
||
MODEL_NAME="moonshot-v1-auto"
|
||
BASE_URL="https://api.moonshot.cn/v1"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="Moonshot API Key"
|
||
KEY_HINT="获取地址: https://platform.moonshot.cn/console/api-keys"
|
||
NEED_KEY=true
|
||
;;
|
||
3)
|
||
MODEL_NAME="qwen-plus"
|
||
BASE_URL="https://dashscope.aliyuncs.com/compatible-mode/v1"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="通义千问 API Key"
|
||
KEY_HINT="获取地址: https://dashscope.console.aliyun.com/apiKey（有免费额度）"
|
||
NEED_KEY=true
|
||
;;
|
||
4)
|
||
MODEL_NAME="glm-4-plus"
|
||
BASE_URL="https://open.bigmodel.cn/api/paas/v4"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="智谱 API Key"
|
||
KEY_HINT="获取地址: https://open.bigmodel.cn/usercenter/apikeys"
|
||
NEED_KEY=true
|
||
;;
|
||
5)
|
||
MODEL_NAME="abab6.5s-chat"
|
||
BASE_URL="https://api.minimax.chat/v1"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="MiniMax API Key"
|
||
KEY_HINT="获取地址: https://platform.minimaxi.com/"
|
||
NEED_KEY=true
|
||
;;
|
||
6)
|
||
MODEL_NAME="doubao-pro-256k"
|
||
BASE_URL="https://ark.cn-beijing.volces.com/api/v3"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="火山引擎 API Key"
|
||
KEY_HINT="获取地址: https://console.volcengine.com/ark"
|
||
NEED_KEY=true
|
||
;;
|
||
7)
|
||
MODEL_NAME="deepseek-ai/DeepSeek-V3"
|
||
BASE_URL="https://api.siliconflow.cn/v1"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="硅基流动 API Key"
|
||
KEY_HINT="获取地址: https://cloud.siliconflow.cn/account/ak"
|
||
NEED_KEY=true
|
||
;;
|
||
8)
|
||
MODEL_NAME="claude-sonnet-4-20250514"
|
||
BASE_URL=""
|
||
PROVIDER="anthropic"
|
||
KEY_LABEL="Anthropic API Key"
|
||
KEY_HINT="获取地址: https://console.anthropic.com/settings/keys（需翻墙）"
|
||
NEED_KEY=true
|
||
;;
|
||
9)
|
||
MODEL_NAME="gpt-4o"
|
||
BASE_URL=""
|
||
PROVIDER="openai"
|
||
KEY_LABEL="OpenAI API Key"
|
||
KEY_HINT="获取地址: https://platform.openai.com/api-keys（需翻墙）"
|
||
NEED_KEY=true
|
||
;;
|
||
10)
|
||
MODEL_NAME="llama3.2"
|
||
BASE_URL="http://127.0.0.1:11434/v1"
|
||
PROVIDER="custom"
|
||
KEY_LABEL=""
|
||
KEY_HINT="先安装 Ollama (https://ollama.com)，然后运行: ollama run llama3.2"
|
||
NEED_KEY=false
|
||
;;
|
||
*)
|
||
echo -e " ${YELLOW}未知选项，使用默认 DeepSeek${NC}"
|
||
MODEL_NAME="deepseek-chat"
|
||
BASE_URL="https://api.deepseek.com/v1"
|
||
PROVIDER="custom"
|
||
KEY_LABEL="DeepSeek API Key"
|
||
KEY_HINT="获取地址: https://platform.deepseek.com/api_keys"
|
||
NEED_KEY=true
|
||
;;
|
||
esac
|
||
|
||
echo ""
|
||
|
||
# 输入 API Key
|
||
API_KEY=""
|
||
if [ "$NEED_KEY" = "true" ]; then
|
||
echo -e " ${CYAN}$KEY_HINT${NC}"
|
||
echo ""
|
||
read -p " 请输入 $KEY_LABEL: " API_KEY
|
||
if [ -z "$API_KEY" ]; then
|
||
echo -e " ${YELLOW}⚠ 未输入 API Key，稍后可通过 Config.html 配置${NC}"
|
||
fi
|
||
else
|
||
echo -e " ${CYAN}$KEY_HINT${NC}"
|
||
fi
|
||
|
||
# 转义 API_KEY 中可能破坏 JSON 的字符 (反斜杠和双引号)。这里只处理这两个，
|
||
# 因为合法的 API key 极少包含控制字符。如果用户粘贴了奇怪的内容,至少 JSON 仍然可解析。
|
||
API_KEY_JSON=${API_KEY//\\/\\\\}
|
||
API_KEY_JSON=${API_KEY_JSON//\"/\\\"}
|
||
|
||
# 写配置文件
|
||
if [ "$PROVIDER" = "custom" ] && [ -n "$BASE_URL" ]; then
|
||
cat > "$CONFIG_PATH" << CFGEOF
|
||
{
|
||
"gateway": {
|
||
"mode": "local",
|
||
"auth": { "token": "uclaw" }
|
||
},
|
||
"models": {
|
||
"mode": "merge",
|
||
"providers": {
|
||
"custom": {
|
||
"baseUrl": "$BASE_URL",
|
||
"apiKey": "$API_KEY_JSON",
|
||
"api": "openai-completions",
|
||
"models": [{ "id": "$MODEL_NAME" }]
|
||
}
|
||
}
|
||
},
|
||
"agents": { "defaults": { "model": { "primary": "custom/$MODEL_NAME" } } }
|
||
}
|
||
CFGEOF
|
||
elif [ "$PROVIDER" = "anthropic" ]; then
|
||
cat > "$CONFIG_PATH" << CFGEOF
|
||
{
|
||
"gateway": {
|
||
"mode": "local",
|
||
"auth": { "token": "uclaw" }
|
||
},
|
||
"models": {
|
||
"mode": "merge",
|
||
"providers": {
|
||
"anthropic": {
|
||
"apiKey": "$API_KEY_JSON",
|
||
"api": "anthropic",
|
||
"models": [{ "id": "$MODEL_NAME" }]
|
||
}
|
||
}
|
||
},
|
||
"agents": { "defaults": { "model": { "primary": "anthropic/$MODEL_NAME" } } }
|
||
}
|
||
CFGEOF
|
||
elif [ "$PROVIDER" = "openai" ]; then
|
||
cat > "$CONFIG_PATH" << CFGEOF
|
||
{
|
||
"gateway": {
|
||
"mode": "local",
|
||
"auth": { "token": "uclaw" }
|
||
},
|
||
"models": {
|
||
"mode": "merge",
|
||
"providers": {
|
||
"openai": {
|
||
"apiKey": "$API_KEY_JSON",
|
||
"api": "openai-completions",
|
||
"models": [{ "id": "$MODEL_NAME" }]
|
||
}
|
||
}
|
||
},
|
||
"agents": { "defaults": { "model": { "primary": "openai/$MODEL_NAME" } } }
|
||
}
|
||
CFGEOF
|
||
fi
|
||
|
||
echo -e " ${GREEN}✓${NC} 模型配置完成: $MODEL_NAME"
|
||
else
|
||
# 非交互模式，写默认配置
|
||
echo -e " ${YELLOW}⚠${NC} 管道模式，跳过模型选择"
|
||
echo -e " ${DIM}启动后用 Config.html 配置模型${NC}"
|
||
if [ ! -f "$CONFIG_PATH" ]; then
|
||
cat > "$CONFIG_PATH" << 'CFGEOF'
|
||
{
|
||
"gateway": {
|
||
"mode": "local",
|
||
"auth": { "token": "uclaw" }
|
||
}
|
||
}
|
||
CFGEOF
|
||
fi
|
||
fi
|
||
fi
|
||
|
||
echo ""
|
||
|
||
# ============================================================
|
||
# Step 7: 生成启动脚本 + 验证 + 摘要
|
||
# ============================================================
|
||
echo -e " ${BOLD}[6/7] 生成启动脚本 ...${NC}"
|
||
|
||
# Mac 启动脚本
|
||
cat > "$UCLAW_DIR/start.command" << 'STARTEOF'
|
||
#!/bin/bash
|
||
DIR="$(cd "$(dirname "$0")" && pwd)"
|
||
ARCH=$(uname -m)
|
||
|
||
# Find Node.js
|
||
if [ "$ARCH" = "arm64" ]; then
|
||
NODE_BIN="$DIR/runtime/node-mac-arm64/bin/node"
|
||
else
|
||
NODE_BIN="$DIR/runtime/node-mac-x64/bin/node"
|
||
fi
|
||
[ ! -f "$NODE_BIN" ] && NODE_BIN="$(which node 2>/dev/null)"
|
||
|
||
if [ -z "$NODE_BIN" ] || [ ! -f "$NODE_BIN" ]; then
|
||
echo "错误: 未找到 Node.js"
|
||
exit 1
|
||
fi
|
||
|
||
OPENCLAW_MJS="$DIR/core/node_modules/openclaw/openclaw.mjs"
|
||
if [ ! -f "$OPENCLAW_MJS" ]; then
|
||
echo "错误: 未找到 OpenClaw"
|
||
exit 1
|
||
fi
|
||
|
||
export OPENCLAW_HOME="$DIR/data"
|
||
export OPENCLAW_STATE_DIR="$DIR/data/.openclaw"
|
||
export OPENCLAW_CONFIG_PATH="$DIR/data/.openclaw/openclaw.json"
|
||
|
||
# Find available port
|
||
PORT=18789
|
||
while lsof -i :$PORT >/dev/null 2>&1; do
|
||
PORT=$((PORT + 1))
|
||
[ $PORT -gt 18799 ] && echo "没有可用端口 (18789-18799)" && exit 1
|
||
done
|
||
|
||
cd "$DIR/core"
|
||
"$NODE_BIN" "$OPENCLAW_MJS" gateway run --allow-unconfigured --force --port $PORT &
|
||
PID=$!
|
||
|
||
# Wait for gateway to start, then open browser
|
||
for i in $(seq 1 30); do
|
||
sleep 0.5
|
||
if curl -s -o /dev/null "http://127.0.0.1:$PORT/" 2>/dev/null; then
|
||
open "http://127.0.0.1:$PORT/#token=uclaw" 2>/dev/null || true
|
||
break
|
||
fi
|
||
done
|
||
|
||
echo ""
|
||
echo " 🦞 U-Claw 正在运行: http://127.0.0.1:$PORT/#token=uclaw"
|
||
echo " 按 Ctrl+C 停止"
|
||
echo ""
|
||
|
||
wait $PID
|
||
STARTEOF
|
||
chmod +x "$UCLAW_DIR/start.command"
|
||
|
||
# Linux 启动脚本
|
||
if [ "$OS" = "Linux" ]; then
|
||
cat > "$UCLAW_DIR/start.sh" << 'STARTEOF'
|
||
#!/bin/bash
|
||
DIR="$(cd "$(dirname "$0")" && pwd)"
|
||
ARCH=$(uname -m)
|
||
|
||
if [ "$ARCH" = "x86_64" ]; then
|
||
NODE_BIN="$DIR/runtime/node-linux-x64/bin/node"
|
||
elif [ "$ARCH" = "aarch64" ]; then
|
||
NODE_BIN="$DIR/runtime/node-linux-arm64/bin/node"
|
||
fi
|
||
[ ! -f "$NODE_BIN" ] && NODE_BIN="$(which node 2>/dev/null)"
|
||
|
||
OPENCLAW_MJS="$DIR/core/node_modules/openclaw/openclaw.mjs"
|
||
|
||
export OPENCLAW_HOME="$DIR/data"
|
||
export OPENCLAW_STATE_DIR="$DIR/data/.openclaw"
|
||
export OPENCLAW_CONFIG_PATH="$DIR/data/.openclaw/openclaw.json"
|
||
|
||
PORT=18789
|
||
while ss -tln | grep -q ":$PORT "; do
|
||
PORT=$((PORT + 1))
|
||
[ $PORT -gt 18799 ] && echo "没有可用端口" && exit 1
|
||
done
|
||
|
||
cd "$DIR/core"
|
||
"$NODE_BIN" "$OPENCLAW_MJS" gateway run --allow-unconfigured --force --port $PORT &
|
||
PID=$!
|
||
|
||
sleep 2
|
||
echo ""
|
||
echo " 🦞 U-Claw 正在运行: http://127.0.0.1:$PORT/#token=uclaw"
|
||
echo " 按 Ctrl+C 停止"
|
||
echo ""
|
||
|
||
xdg-open "http://127.0.0.1:$PORT/#token=uclaw" 2>/dev/null || true
|
||
wait $PID
|
||
STARTEOF
|
||
chmod +x "$UCLAW_DIR/start.sh"
|
||
fi
|
||
|
||
echo -e " ${GREEN}✓${NC} 启动脚本已生成"
|
||
echo ""
|
||
|
||
# ============================================================
|
||
# 验证
|
||
# ============================================================
|
||
echo -e " ${BOLD}[7/7] 验证安装 ...${NC}"
|
||
echo ""
|
||
|
||
# Node.js
|
||
if [ -n "$INSTALL_NODE" ] && "$INSTALL_NODE" --version >/dev/null 2>&1; then
|
||
NODE_VER_STR=$("$INSTALL_NODE" --version)
|
||
echo -e " ${GREEN}[✓]${NC} Node.js $NODE_VER_STR"
|
||
else
|
||
echo -e " ${RED}[✗]${NC} Node.js"
|
||
fi
|
||
|
||
# OpenClaw
|
||
if [ -f "$CORE_DIR/node_modules/openclaw/openclaw.mjs" ]; then
|
||
echo -e " ${GREEN}[✓]${NC} OpenClaw 已安装"
|
||
else
|
||
echo -e " ${RED}[✗]${NC} OpenClaw"
|
||
fi
|
||
|
||
# QQ 插件
|
||
if [ -d "$CORE_DIR/node_modules/@sliverp/qqbot" ]; then
|
||
echo -e " ${GREEN}[✓]${NC} QQ 插件"
|
||
else
|
||
echo -e " ${YELLOW}[⚠]${NC} QQ 插件（未安装，不影响主功能）"
|
||
fi
|
||
|
||
# 技能
|
||
INSTALLED_SKILLS=$(ls -d "$SKILLS_TARGET"/*/ 2>/dev/null | wc -l | tr -d ' ')
|
||
echo -e " ${GREEN}[✓]${NC} 中国技能 (${INSTALLED_SKILLS}个)"
|
||
|
||
# 配置文件
|
||
if [ -f "$CONFIG_PATH" ]; then
|
||
echo -e " ${GREEN}[✓]${NC} 配置文件"
|
||
else
|
||
echo -e " ${YELLOW}[⚠]${NC} 配置文件（需启动后配置）"
|
||
fi
|
||
|
||
echo ""
|
||
|
||
# ============================================================
|
||
# 摘要
|
||
# ============================================================
|
||
INSTALL_SIZE=$(du -sh "$UCLAW_DIR" 2>/dev/null | cut -f1)
|
||
|
||
echo -e "${GREEN}${BOLD}"
|
||
echo " ╔══════════════════════════════════════════╗"
|
||
echo " ║ ✅ U-Claw 安装成功！ ║"
|
||
echo " ╚══════════════════════════════════════════╝"
|
||
echo -e "${NC}"
|
||
echo -e " ${BOLD}安装位置:${NC} $UCLAW_DIR"
|
||
echo -e " ${BOLD}大小:${NC} $INSTALL_SIZE"
|
||
echo ""
|
||
echo -e " ${BOLD}启动方式:${NC}"
|
||
if [ "$OS" = "Darwin" ]; then
|
||
echo -e " 双击 ${CYAN}~/.uclaw/start.command${NC}"
|
||
echo -e " 或终端: ${CYAN}bash ~/.uclaw/start.command${NC}"
|
||
else
|
||
echo -e " 终端: ${CYAN}bash ~/.uclaw/start.sh${NC}"
|
||
fi
|
||
echo ""
|
||
echo -e " ${BOLD}打开后:${NC}"
|
||
echo -e " 浏览器自动打开 → 开始和 AI 对话"
|
||
echo ""
|
||
echo -e " ${DIM}如需重新配置模型，编辑 ~/.uclaw/data/.openclaw/openclaw.json${NC}"
|
||
echo -e " ${DIM}卸载: rm -rf ~/.uclaw${NC}"
|
||
echo ""
|