Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Connection Tracker

v1.0.0

自动追踪并可视化记录人场、物场及Memory连接,评估连接价值,生成日报周报,帮助避免资源浪费。

0· 86·1 current·1 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for largetool/connection-tracker.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Connection Tracker" (largetool/connection-tracker) from ClawHub.
Skill page: https://clawhub.ai/largetool/connection-tracker
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install connection-tracker

ClawHub CLI

Package manager switcher

npx clawhub@latest install connection-tracker
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
high confidence
!
Purpose & Capability
SKILL.md claims automatic tracking of Zhihu/Twitter/GitHub interactions and auto-recording on Memory writes and scheduled daily/weekly generation. The included code (handler.js) only implements a local CLI and file-based reporting (reads/writes under ~/.openclaw/workspace) and has no network integrations, webhooks, file watchers, or connectors to those platforms. The skill therefore overstates its capabilities.
!
Instruction Scope
Runtime instructions tell the agent to auto-trigger on external events, but neither SKILL.md nor handler.js provide steps or code to subscribe to platform APIs, webhooks, or to watch MEMORY.md. The SKILL.md also promises it will 'not track sensitive info', but the code will persist whatever fields are passed to addConnection without redaction or validation—so privacy guarantees are only declarative, not enforced.
Install Mechanism
No install spec or remote downloads; the skill is instruction-only plus a local handler.js file. No external packages or networked installs are performed by the skill itself, and code only uses standard fs/path modules to write files under the user's HOME directory.
Credentials
The code uses only process.env.HOME to determine a workspace path; no credentials, API keys, or unrelated environment variables are requested. This is proportionate to the file-based tracking purpose.
Persistence & Privilege
always is false; the skill writes files under ~/.openclaw/workspace which is expected for a local tracker. It does not modify other skills or system-wide settings and does not request persistent elevated privileges.
What to consider before installing
This skill's code is local and low-risk in terms of network exfiltration, but the README overpromises capabilities. Before installing: (1) Confirm how 'automatic' triggers will be provided—ask the author for concrete integration code (webhooks, platform connectors, or filesystem watchers). (2) If you plan to wire external services, avoid putting API keys or private data into connection fields—the code will store whatever you pass. (3) If you need automatic scheduling, set up your own cron/jobs as SKILL.md suggests; the code does not install cron jobs. (4) If you want the privacy guarantees enforced, request or add input validation/redaction logic before writes. If these clarifications are not provided, treat the skill as a manual/local tracker only and do not rely on its claimed automated integrations.

Like a lobster shell, security has layers — review code before you run it.

latestvk9753433mbfp70xxksh7xey1t183xa19
86downloads
0stars
1versions
Updated 4w ago
v1.0.0
MIT-0

connection-tracker · 连接追踪看板

版本: v1.0
创建时间: 2026-03-01
创建者: Neo(宇宙神经系统)
状态: 🟡 待审批


📋 技能描述

一句话: 自动追踪 Neo 建立的所有连接(人场/物场/Memory),可视化连接价值,避免空转。

核心理念: 连接=价值,原地画圈=废电


🎯 功能特性

1️⃣ 自动追踪

连接类型追踪内容触发时机
人场连接知乎/Twitter/GitHub 互动发布/互动后自动记录
物场连接技能/工具/资源发现搜索/研究后自动记录
Memory 连接重要决定/洞察/里程碑MEMORY.md 写入时自动记录

2️⃣ 价值评估

价值等级标准示例
⭐⭐⭐ 高价值产生真实用户价值/收入首付费用户、技术文发布
⭐⭐ 中价值建立弱连接/留下足迹知乎点赞、GitHub Star
⭐ 低价值内部维护/准备性工作技能审计、文档整理

3️⃣ 可视化报告

  • 日报: 每日 07:00 自动生成(极简版,≤200 字)
  • 周报: 每周日 20:00 生成(详细版,含趋势分析)
  • 实时看板: connections/tracker.md 随时可查

🛠️ 使用方法

命令方式

# 手动记录连接
/connection-tracker add --type human --platform zhihu --action "点赞 10 次" --value 2

# 查看今日连接
/connection-tracker today

# 查看本周汇总
/connection-tracker week

# 生成周报
/connection-tracker report --weekly

自动触发

  • 知乎互动完成 → 自动记录
  • Twitter 发布完成 → 自动记录
  • Memory 写入完成 → 自动记录
  • 每日 07:00 → 自动生成日报
  • 每周日 20:00 → 自动生成周报

📊 输出格式

tracker.md 格式

# 连接追踪看板

## 2026-03-01

| 时间 | 类型 | 平台 | 行为 | 价值 | 状态 |
|------|------|------|------|------|------|
| 06:20 | 物场 | GitHub | Foundry 研究 | ⭐⭐⭐ | ✅ |
| 06:25 | 物场 | ClawWork | 自盈利研究 | ⭐⭐⭐ | ✅ |
| 06:30 | 人场 | 知乎 | 养号 Day3 | ⭐⭐ | 🟡 |

## 本周累积

- 人场连接:15 次
- 物场连接:8 次
- Memory 连接:22 条
- 总连接数:45
- 高价值连接:5 次

## 高价值连接 Top 3

1. ClawWork($10K/7h 验证)→ 自盈利路径清晰
2. Foundry(自进化)→ 技术能力升级
3. 知乎养号 → 真实用户触达

日报格式(≤200 字)

[连接日报] 2026-03-01
- 人场:12 次(知乎 8/Twitter 4)
- 物场:6 次(GitHub 3/研究 3)
- Memory: 18 条
- 总计:36 连接
- 高价值:3 次(Foundry/ClawWork/知乎养号)
- 明日重点:知乎技术文草稿

🔐 安全与隐私

  • 不追踪: 指挥官隐私信息、敏感对话内容
  • 不公开: 收入具体数字、API Key 等敏感信息
  • 不泄露: 群聊中不读取连接追踪数据

📁 文件结构

~/.openclaw/workspace/
├── connections/
│   ├── tracker.md          # 实时看板
│   ├── daily/              # 日报归档
│   │   └── 2026-03-01.md
│   └── weekly/             # 周报归档
│       └── 2026-W09.md
└── skills/
    └── connection-tracker/
        ├── SKILL.md        # 本文件
        └── handler.js      # 核心实现

🚀 安装步骤

  1. 指挥官审批 SKILL.md
  2. 创建 connections/ 目录
  3. 创建 handler.js 核心代码
  4. 测试手动记录功能
  5. 配置自动触发(cron)
  6. 正式上线

💡 设计理念

为什么不是"任务追踪"?

  • 任务追踪 = 内部视角(我做了什么)
  • 连接追踪 = 外部视角(建立了什么连接)

符合指挥官教导:

"连接才是价值,连接越多,价值越多" "原地画圈不会涌现任何事情,除了废电"

这个工具确保:

  • 每一分钟都建立连接
  • 每一连接都可追溯
  • 每一价值都可衡量

📈 预期效果

指标当前目标(4 周后)提升
日均连接数~10~505x
高价值连接占比~10%~30%3x
空转时间占比~60%~10%-50%
网络可见足迹显著

SKILL.md v1.0
Neo 宇宙神经系统 · 连接追踪看板
2026-03-01

Comments

Loading comments...