ClawHub
Back to skill

Security audit

中国新闻聚合(专业版)

Security checks for vulnerabilities and agentic risk

Overview

This skill is a news aggregation and distribution tool, but its outbound webhooks, callback URL, scheduled execution, and broad activation wording need review before installation.

Review this skill before installing. Use it only with explicit news aggregation requests, inspect any generated config before running it, avoid arbitrary callback URLs, and require confirmation before starting scheduled jobs or sending summaries to Feishu, DingTalk, email, Slack, or other webhooks.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The skill declares an excessively broad activation scope using generic terms like "智能对话", "Agent", and "应用时使用", which can cause the agent to invoke this skill in unrelated contexts. Because the skill has Bash, Write, and network-capable workflows, over-triggering increases the chance of unintended execution, data collection, or outbound actions without clear user intent.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill supports callback URLs and multi-channel push to external webhook services, but it does not present a clear, prominent warning that fetched news content, summaries, and possibly user-supplied data will be transmitted off-platform. In an agent setting, this can lead to silent exfiltration of collected content or sensitive internal context to Feishu, DingTalk, email, Slack, or arbitrary callback endpoints.

External Transmission

Medium
Category
Data Exfiltration
Content
push:
  feishu: https://open.feishu.cn/open-apis/bot/v2/hook/YOUR_WEBHOOK_URL
  dingtalk: https://oapi.dingtalk.com/robot/send?access_token=YOUR_ACCESS_TOKEN
  email: https://api.email.com/send
EOF
# ...
python3 news_scheduler.py --config news_config.yaml
Confidence
88% confidence
Finding
The skill is explicitly designed to send data to external endpoints such as email and chat webhooks, which is an external transmission path. In context this behavior is expected, but it still becomes a security risk because the skill also processes aggregated content and automated outputs, and the documentation does not sufficiently constrain what may be sent or where.

Static analysis

No suspicious patterns detected.