Skill flagged — suspicious patterns detected

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

Skill Curator

v1.0.0

OpenClaw 的总调度 Skill。当用户在 Discord 发送「【关键词】URL」格式的消息时,自动触发本 Skill:判断该关键词对应的 Skill 是否已存在 → 提取 URL 内容 → 追加或新建知识 Skill → 推送到 GitHub skillhub 仓库。

0· 76·0 current·0 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 eeyan2025-art/skill-curator.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Skill Curator" (eeyan2025-art/skill-curator) from ClawHub.
Skill page: https://clawhub.ai/eeyan2025-art/skill-curator
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 skill-curator

ClawHub CLI

Package manager switcher

npx clawhub@latest install skill-curator
Security Scan
Capability signals
Requires OAuth token
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The SKILL.md describes cloning a GitHub repo and pushing changes and requiring GITHUB_TOKEN and MINIMAX_API_KEY, but the registry metadata lists no required environment variables or primary credential. The skill also hardcodes a default repo (github.com/eeyan2025-art/skillhub.git), which is unexpected if a user expects changes to go to their own repo.
!
Instruction Scope
Runtime instructions tell the agent to fetch arbitrary URLs, extract content (including audio/video analysis), create or modify SKILL.md files, and push commits to a remote GitHub repo. Extracting arbitrary remote content and then pushing it to a repo controlled by another account (or unspecified repo) broadens the scope beyond simple local processing and can cause unwanted data exfiltration or copyright/consent issues.
Install Mechanism
Instruction-only skill with no install spec or bundled code, so nothing is written to disk by an installer. Lowest installation risk from packaging/hosts perspective.
!
Credentials
SKILL.md requires GITHUB_TOKEN (repo push rights) and MINIMAX_API_KEY, but the skill metadata declared no required env vars. Requesting a token with push access is proportionate to pushing changes, but the hardcoded external repo and missing metadata declaration are mismatches. A broad-scope GITHUB_TOKEN would let the skill modify any repos the token can access—this is sensitive and should be scoped and confirmed by the user.
Persistence & Privilege
Skill is not flagged always:true and has no install-time persistence. Autonomous invocation is allowed by default; combined with a push-capable GITHUB_TOKEN this increases risk (agent could autonomously push changes), but autonomous invocation alone is normal.
What to consider before installing
Before installing or enabling this skill: 1) Confirm which GitHub repository it will push to — do not provide a token with broad repo access; prefer a token scoped to a single test repo you control or use a deploy key. 2) Verify why registry metadata omits required env vars (GITHUB_TOKEN, MINIMAX_API_KEY) and ask the publisher to correct metadata. 3) Understand that the skill will fetch arbitrary URLs and may commit extracted content; avoid giving it access to sensitive data and consider requiring explicit user confirmation before any git push. 4) If you must test, use a throwaway GitHub account/repo and limited-scope token, and review generated SKILL.md locally before allowing pushes to production repos. 5) If the hardcoded default repo is not yours, do not provide push credentials until the repo target is changed to a repository you control.

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

latestvk97bpanb9kd5g1d0knsfhg7gfh84esf1
76downloads
0stars
1versions
Updated 2w ago
v1.0.0
MIT-0

Skill Curator — 知识策展人

将任意来源的内容策展为可积累、可迭代的私人 Skill 知识库。

触发条件

用户发送格式:【关键词】URL

示例:

【python】https://bilibili.com/video/BVxxx
【写作技巧】https://mp.weixin.qq.com/s/xxx
【AI绘画】https://xiaohongshu.com/explore/xxx

完整工作流程

Discord 消息: 【关键词】URL
       │
       ▼
┌─────────────────────────┐
│ 1. parse_input          │ ← 解析关键词 + URL
│    keyword: "python"    │
│    url: "https://..."    │
└──────────┬──────────────┘
           ▼
┌─────────────────────────┐
│ 2. clone_repo            │ ← 克隆 GitHub 仓库到本地
│    github.com/           │
│    eeyan2025-art/        │
│    skillhub.git          │
└──────────┬──────────────┘
           ▼
┌─────────────────────────┐
│ 3. check_existing_skill  │ ← 判断 skill 文件是否已存在
│    skills/python/       │
│      └── SKILL.md        │
└──────────┬──────────────┘
    │ 存在       │ 不存在
    ▼            ▼
┌─────────┐  ┌──────────────────┐
│ 追加模式 │  │ 新建模式          │
└────┬────┘  └────────┬─────────┘
     │                 │
     ▼                 ▼
┌─────────────────────────┐
│ 4. extract_content      │ ← 根据 URL 类型提取内容
│   • 视频 → MiniMax 音视频 │
│   • 文章 → 网页抓取       │
│   • 输出:原文 + 摘要     │
└──────────┬──────────────┘
           ▼
┌─────────────────────────┐
│ 5. generate_skill       │ ← 生成/更新 SKILL.md
│   新建:完整 SKILL.md    │
│   追加:追加到现有文件    │
└──────────┬──────────────┘
           ▼
┌─────────────────────────┐
│ 6. git_push             │ ← 提交推送 GitHub
└──────────┬──────────────┘
           ▼
        完成!

输入格式

用户消息必须包含:

  • 【】 包裹的关键词(英文或中文均可)
  • 一个有效的 URL

URL 类型与处理方式

URL 类型处理方式工具
YouTubeMiniMax 音视频分析videos_understand
Bilibili网页提取字幕 + 音频分析extract_content_from_websites + audios_understand
西瓜视频音频分析audios_understand
微信公众号网页提取正文extract_content_from_websites
小红书音频分析audios_understand
任意网页提取正文extract_content_from_websites

追加模式规则(已有 Skill 时)

追加内容到现有 SKILL.md 时:

  1. 不要破坏现有结构(frontmatter、章节结构保持不变)
  2. 在正文末尾追加新内容,格式:
---

## 【YYYY-MM-DD】新增:<来源标题>

<内容摘要>

来源:<URL>
  1. 如果现有 Skill 已包含相同内容(重复检测),跳过不重复添加

GitHub 仓库配置

默认仓库:https://github.com/eeyan2025-art/skillhub.git 默认分支:main Skill 存放路径:skills/<keyword>/SKILL.md

环境变量

需要设置:

  • GITHUB_TOKEN:GitHub Personal Access Token(拥有 repo push 权限)
  • MINIMAX_API_KEY:MiniMax API Key(用于音视频分析)

错误处理

情况处理
GitHub 仓库无对应 Skill新建一个
URL 无法访问尝试降级方案或提示用户
内容提取失败回复用户说明情况,附上可手动处理的方式
Git push 失败输出本地文件路径供手动处理

长期积累机制

每个关键词的 Skill 都是一个持续生长的知识文件,随时间越积越多。比如:

skills/
├── python/
│   └── SKILL.md   ← 不断追加,越来越大
├── 写作技巧/
│   └── SKILL.md
└── ai绘画/
    └── SKILL.md

用户无需每次指定"新建",系统自动判断,自动迭代。

Comments

Loading comments...