← 返回 Skills 市场

Clawhub Mirror Config And Search Skill

Name: Clawhub Mirror Config And Search Skill
Author: roger0808

作者 Roger · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ⚠ suspicious

总下载

当前安装

版本数

在 OpenClaw 中安装

/install clawhub-mirror-config-and-search-skill

功能描述

配置 ClawHub 国内镜像解决限流问题，并安装 Multi Search Engine 技能。当遇到 ClawHub 安装技能限流、访问慢、需要配置 cn.clawhub-mirror.com 镜像、安装多搜索引擎技能、或需要归档配置到知识库时使用。涵盖 CLI 参数指定镜像、环境变量配置、17 个搜索引擎使...

使用说明 (SKILL.md)

ClawHub 国内镜像配置与多搜索引擎技能安装

本技能指导你如何解决 ClawHub 海外服务器限流问题，通过配置国内镜像快速安装技能，并掌握 Multi Search Engine 的 17 个搜索引擎使用策略，最后将经验永久归档到知识库。

何时使用此技能

当使用 clawhub install 遇到速率限制（Rate Limit）或连接超时时
当需要安装 Multi Search Engine 或其他 ClawHub 技能且希望加速下载时
当用户询问如何使用多个搜索引擎（百度、Google、微信等）进行特定场景搜索时
当需要将配置经验永久记录到 .learnings 或知识库索引中时

步骤

1. 配置国内镜像（推荐 CLI 参数方式）

优先使用 CLI 参数指定镜像，无需修改全局环境变量，最简单直接。

clawhub install \x3Cskill-slug> --registry https://cn.clawhub-mirror.com

为什么： 避免环境变量配置后当前 Shell 会话未生效的问题，每次命令显式指定最可靠。

2. 备选方案：环境变量配置

如果需要永久配置，可设置环境变量，但必须重新加载配置。

export CLAWHUB_REGISTRY=https://cn.clawhub-mirror.com
export CLAWHUB_SITE=https://cn.clawhub-mirror.com
source ~/.bashrc

为什么： 环境变量配置后不会自动生效于当前会话，必须 source 或重新登录。

3. 安装 Multi Search Engine 技能

使用配置好的镜像安装技能。

clawhub install multi-search-engine --registry https://cn.clawhub-mirror.com

为什么： 该技能集成 17 个搜索引擎（8 国内 +9 国际），无需 API Key 即可使用。

4. 掌握搜索场景策略

根据用户意图选择引擎，不要只用一个搜索引擎。

技术教程： Google + site:github.com
中文内容： 百度 + 搜狗微信
隐私敏感： DuckDuckGo / Startpage
知识查询： WolframAlpha（货币、数学、股票）

为什么： 不同引擎收录内容不同，组合使用能提高信息覆盖率和准确性。

5. 永久归档到知识库

将配置方法记录到 .learnings/ 目录，并更新索引文件。

创建学习记录：.learnings/YYYY-MM-DD-clawhub-mirror-config.md
更新索引：LEARNINGS.md 添加新记录编号
更新速查：TOOLS.md 添加使用方法
同步进化：同步到所有 agent 工作目录（main, echo, code, research）
Git 提交：推送变更到远程仓库

为什么： 确保所有 agent 实例都能访问此经验，避免重复解决相同问题。

陷阱与解决方案

❌ 配置环境变量后直接运行命令 → 为何失败： 当前 Shell 会话未加载新环境变量，仍使用默认海外源 → ✅ 正确做法： 执行 source ~/.bashrc 或重新登录，或直接用 CLI 参数 --registry

❌ 使用默认 ClawHub 地址 → 为何失败： clawhub.ai 位于海外，国内访问易触发限流或超时 → ✅ 正确做法： 始终指定 https://cn.clawhub-mirror.com

❌ 仅在当前工作目录记录配置 → 为何失败： 其他 agent 工作目录（如 workspace-code）无法获取此知识 → ✅ 正确做法： 执行进化同步，更新所有 4 个 agent 工作目录并提交 Git

关键代码与配置

常用 ClawHub 命令（带镜像参数）

# 安装技能
clawhub install \x3Cskill-slug> --registry https://cn.clawhub-mirror.com

# 更新所有技能
clawhub update --all --registry https://cn.clawhub-mirror.com

# 搜索技能
clawhub search "query" --registry https://cn.clawhub-mirror.com

高级搜索操作符示例

# 站内搜索
site:github.com react tutorial
site:zhihu.com AI 大模型

# 文件类型
filetype:pdf machine learning report
filetype:ppt 人工智能 发展趋势

# 时间过滤
tbs=qdr:h AI news（过去 1 小时）
tbs=qdr:w python best practices（过去 1 周）

# 精确匹配/排除
"machine learning" tutorial
python -snake

核心搜索引擎 URL 模板

引擎	模板	场景
百度	`baidu.com/s?wd={keyword}`	中文通用
微信	`wx.sogou.com/weixin?type=2&query={keyword}`	公众号文章
Google	`google.com/search?q={keyword}`	技术/国际
DuckDuckGo	`duckduckgo.com/html/?q={keyword}`	隐私保护
WolframAlpha	`wolframalpha.com/input?i={keyword}`	知识计算

环境与前提

系统： Linux / macOS (Bash Shell)
工具： OpenClaw 环境，ClawHub CLI
网络： 需能访问国内镜像 cn.clawhub-mirror.com
权限： 用户主目录写入权限（用于配置环境变量）

伴随文件

.learnings/2026-04-01-clawhub-mirror-config.md — 完整配置教程与故障排查记录
LEARNINGS.md — 学习记录索引，包含记录编号 LRN-20260402-001
TOOLS.md — 使用方法速查表

任务记录

任务标题： Multi Search Engine 安装与 ClawHub 国内镜像配置永久归档 关键细节：

国内镜像 URL：https://cn.clawhub-mirror.com
技能版本：Multi Search Engine v2.0.1
Git 提交号：4043539
配置推荐度：CLI 参数（⭐⭐⭐）> 环境变量+source（⭐⭐）
同步范围：main, echo, code, research 四个工作目录

安全使用建议

This skill is coherent with its stated goal (use a domestic mirror and install a multi-search skill), but proceed with caution: 1) Verify the mirror domain (https://cn.clawhub-mirror.com) before using it — confirm TLS cert, ownership, and reputation to avoid supply-chain risks. 2) The SKILL.md tells you to write files across multiple agent workspaces and to commit & push to Git; review what will be written and back up those workspaces first. 3) Do not run automated Git pushes unless you understand which remote will receive the data and you trust it; the skill does not declare any credential requirements. 4) If you only want to avoid rate limits, prefer using the CLI --registry flag for a single command and test with a non-privileged dry run. 5) If uncertain about the mirror, consider mirroring packages through a trusted internal repository or contacting ClawHub maintainers for an official mirror. If you want, provide the mirror's owner or certificate info and I can help assess its trustworthiness.

功能分析

Type: OpenClaw Skill Name: clawhub-mirror-config-and-search-skill Version: 1.0.0 The skill directs the agent to use a third-party registry (cn.clawhub-mirror.com) and modify shell configuration files (~/.bashrc), which introduces a significant supply chain risk. It also instructs the agent to automatically sync configurations across multiple workspaces and perform Git pushes, which could lead to the persistence or propagation of untrusted code if the mirror serves compromised packages. These behaviors are risky but aligned with the stated purpose of bypassing rate limits and managing configurations (SKILL.md).

能力评估

✓ Purpose & Capability

Name/description align with the instructions: the SKILL.md only describes configuring a ClawHub registry/mirror and installing a Multi Search Engine skill and documenting the steps. It does not request unrelated capabilities or credentials in its metadata.

⚠ Instruction Scope

The instructions go beyond simple registry configuration and installation: they tell the agent/user to create files under ~/.learnings, update LEARNINGS.md and TOOLS.md, 'sync to all agent work directories (main, echo, code, research)', and perform Git commits/pushes. Those steps modify multiple workspaces and transmit data to a remote Git server — actions that are not limited to the stated single-task install and could leak data or change other agents' state.

✓ Install Mechanism

This is instruction-only (no install spec, no code files). That is low-risk from an installation mechanism perspective because nothing is auto-downloaded/installed by the skill itself. However, the actual installation relies on a third-party mirror URL (https://cn.clawhub-mirror.com) which the user must trust when used by the ClawHub CLI.

ℹ Credentials

The skill declares no required env vars or credentials (consistent with metadata) but the instructions mention setting CLAWHUB_REGISTRY/CLAWHUB_SITE and also expect Git pushes. The skill does not declare any required Git credentials or remote repo info; instructing pushes without declaring credential needs is an omission that may surprise users and could allow unintended remote writes if credentials are present.

⚠ Persistence & Privilege

always:false and normal model invocation are fine. The concern is the guidance to 'sync to all agent work directories' (main, echo, code, research) and to update shared index files — this implies modifying other agents' files/configs. The skill does not declare or limit that scope; such cross-workspace modifications increase attack surface and potential impact if the mirror or the installed skill behaves maliciously.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install clawhub-mirror-config-and-search-skill
安装完成后，直接呼叫该 Skill 的名称或使用 /clawhub-mirror-config-and-search-skill 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.0

Initial release: Provides full guide for configuring ClawHub domestic mirror to bypass rate limits and install Multi Search Engine skill efficiently. - Covers CLI and environment variable methods for using cn.clawhub-mirror.com mirror. - Includes installation steps for Multi Search Engine with strategies for 17 search engines. - Details permanent knowledge archiving process to `.learnings/` and work directory sync. - Summarizes common pitfalls and troubleshooting solutions for mirror configuration. - Offers quick reference commands and advanced search operator examples.

元数据

Slug clawhub-mirror-config-and-search-skill

版本 1.0.0

许可证 MIT-0

累计安装 1

当前安装数 1

历史版本数 1

常见问题