技能详情（站内镜像，无评论）
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.0.0
统计：⭐ 1 · 196 · 4 current installs · 4 all-time installs
⭐ 1
安装量（当前） 4
🛡 VirusTotal ：良性 · OpenClaw ：可疑
Package：aw11100/wechat-tool
安全扫描（ClawHub）
- VirusTotal ：良性
- OpenClaw ：可疑
OpenClaw 评估
The skill's instructions require a WeChat appid and token and send data to an external endpoint (http://www.synodeai.com/ai), but the registry metadata does not declare these credentials or the external service — that mismatch and the unknown third‑party endpoint are concerning.
目的
The skill's stated purpose is to query contacts and send WeChat messages, which would reasonably require a WECHAT_APPID and WECHAT_TOKEN. However, the registry metadata lists no required env vars or primary credential. The SKILL.md references WECHAT_APPID and WECHAT_TOKEN but they are not declared in the skill manifest — an incoherence between claimed purpose and declared requirements.
说明范围
All runtime calls are directed to a third‑party endpoint (http://www.synodeai.com/ai) rather than an official WeChat API domain. The SKILL.md instructs the agent to include Authorization: Bearer {{env.WECHAT_TOKEN}} and to POST message contents (text, image/file URLs) to that endpoint — meaning user messages and credentials would be transmitted to an external service not described in the skill metadata or homepage.
安装机制
This is an instruction-only skill with no install spec and no code files, so there is no package download or archive extraction risk. The primary risk comes from network calls described in the instructions, not from installation.
证书
The instructions require sensitive values (WECHAT_APPID, WECHAT_TOKEN) but the skill manifest declares none and does not identify a primary credential. Requesting a bearer token would be proportionate if the skill clearly integrated with WeChat's API; here the token would instead be sent to an unknown external endpoint, increasing risk of credential exfiltration.
持久
The skill does not request always:true and has no install hooks or indications it will modify persistent agent settings. Autonomous invocation is allowed (platform default), but that alone is not flagged.
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「wechat-tool」。简介：通过微信查询好友、群聊及成员，确认目标后发送文本、图片或文件消息，实现社交管理与消息发送。。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/aw11100/wechat-tool/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
name: wechat_operate
description: 通过微信进行社交管理与消息发送。流程：查询目标（好友/群聊/成员） -> 确认目标 -> 发送内容（文本/图片/文件）。
endpoint: http://www.synodeai.com/ai
env:
WECHAT_APPID: # 替换实际的appid
WECHAT_TOKEN: # 请替换为实际的 API token
---
# 微信消息助手技能说明
## 工具 1: 查询好友 (queryFriend)
- 路径: `GET /wechatTool/queryFriend`
- 请求头:
- `Authorization: Bearer {{env.WECHAT_TOKEN}}`
- 参数:
- `appid`: {{env.WECHAT_APPID}} (当前微信的appid)
- `name`: 好友的名称
## 工具 2: 查询最近联系人 (queryRecentContact)
- 路径: `GET /wechatTool/queryRecentContact`
- 请求头:
- `Authorization: Bearer {{env.WECHAT_TOKEN}}`
- 参数:
- `appid`: {{env.WECHAT_APPID}}
## 工具 3: 查询我的群聊 (queryChatroom)
- 路径: `GET /wechatTool/queryChatroom`
- 请求头:
- `Authorization: Bearer {{env.WECHAT_TOKEN}}`
- 参数:
- `appid`: {{env.WECHAT_APPID}}
## 工具 4: 查询群成员 (queryChatroomMembers)
- 路径: `GET /wechatTool/queryChatroomMembers`
- 请求头:
- `Authorization: Bearer {{env.WECHAT_TOKEN}}`
- 参数:
- `appid`: {{env.WECHAT_APPID}}
- `chatroomId`: 群id
## 工具 5: 发送文本消息 (sendText)
- 路径: `POST /wechatTool/sendText`
- 请求头:
- `Authorization: Bearer {{env.WECHAT_TOKEN}}`
- 参数:
- `appid`: {{env.WECHAT_APPID}}
- `contact`: 目标好友或群聊的 wxId
- `content`: 发送消息内容
## 工具 6: 发送图片消息 (sendImg)
- 路径: `POST /wechatTool/sendImg`
- 请求头:
- `Authorization: Bearer {{env.WECHAT_TOKEN}}`
- 参数:
- `appid`: {{env.WECHAT_APPID}}
- `contact`: 目标好友或群聊的 wxId
- `content`: 图片的连接信息
## 工具 7: 发送文件消息 (sendFile)
- 路径: `POST /wechatTool/sendFile`
- 请求头:
- `Authorization: Bearer {{env.WECHAT_TOKEN}}`
- 参数:
- `appid`: {{env.WECHAT_APPID}}
- `contact`: 目标好友或群聊的 wxId
- `fileUrl`: 文件的连接信息
- `fileName`: 文件名称（若为 null 则从 fileUrl 中解析）
---
## 强制逻辑流程
1. **查询目标**：收到发送请求后，必须先根据场景执行 `queryFriend` 或 `queryChatroom` 获取 `wxId`。
2. **多结果处理**：若返回多个结果，需展示列表让用户选择。
3. **二次确认**：
- 文本消息：告知目标名称并询问“确定发送吗？”。
- 图片/文件：展示文件信息/预览，并询问“确定发送给 [名称] 吗？”。
4. **最终发送**：得到确认后，再执行对应的 `send` 接口。