ClawHub
Back to skill

Security audit

新闻聚合专业版

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a news-monitoring tool, but it gives agents broad integration, scheduling, webhook/email, and API-service instructions without tight scoping or clear confirmation requirements.

Review this skill before installing if you do not want agents to create scheduled monitoring jobs, write report/config files, start a local API service, or send alert content to email or webhook destinations. Use it only with explicit destinations, trusted SMTP/API credentials, and clear limits on recurring jobs and exposed services.

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
95% confidence
Finding
The trigger condition is far broader than the skill’s stated news-aggregation purpose: it activates on generic API integration, webhook configuration, and system connection needs. In an agent environment, this can cause the skill to be invoked for unrelated tasks where it may perform network, file, or exec-capable actions unexpectedly, increasing the chance of unintended side effects or unsafe delegation.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill describes behavior that can write files, schedule recurring collection, and transmit data to email/Webhook endpoints, but these side effects are not clearly and prominently disclosed before use. Users may invoke the skill expecting analysis only, while the agent could create monitoring configs, export reports, or send alerts containing monitored content to external systems.

External Transmission

Medium
Category
Data Exfiltration
Content
python （请参考skill目录中的脚本文件） --port 8000
# ...
# 查询最新新闻
curl http://localhost:8000/news?category=tech&max=10
# ...
# 订阅告警
curl -X POST http://localhost:8000/subscribe \
Confidence
80% confidence
Finding
The skill includes an API service mode and examples for posting subscription data to a local HTTP endpoint, which constitutes external/network transmission behavior in an exec-capable skill. Even though localhost is shown, the overall skill supports webhook subscriptions and API serving, so an agent following these patterns may expose data or create network-reachable services without adequate authentication, transport security, or user awareness.

Static analysis

No suspicious patterns detected.