ClawHub
Back to skill

Security audit

daily-news-push

Security checks across malware telemetry and agentic risk

Overview

The skill fits a daily news push use case, but it needs review because setup can write raw user input into executable Python config while the workflow sends full reports to external channels.

Install only if you are comfortable reviewing the generated config.py before running it. Use trusted webhook endpoints, avoid pasting header snippets from untrusted sources, protect OAuth and Authorization tokens, expect the full report to be sent to the configured destination, and add cron only when you want ongoing unattended sends.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (5)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill documentation describes scripts that generate configuration files and send content to external services, implying file-write and network capabilities, but no explicit permissions are declared. This creates a trust and consent gap: users or platforms may invoke a skill with side effects they were not clearly warned about, including writing local config and sending data off-host.

Tp4

High
Category
MCP Tool Poisoning
Confidence
89% confidence
Finding
The documented behavior overstates automation and may cause users to trust the skill as a complete news-collection service when it is primarily a push framework with real outbound webhook capability. That mismatch is dangerous because users may authorize external delivery assuming content provenance and filtering are implemented, when in reality unverified or empty content may be sent to third-party endpoints.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The description does not prominently warn users that generated content may be automatically pushed to external messaging platforms or arbitrary webhooks. This is risky because users may provide sensitive topics, summaries, or links without realizing the skill will transmit them outside the local environment.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The Webhook section explicitly documents sending the full report body to an arbitrary external URL and shows optional Authorization headers, but it does not warn users that potentially sensitive generated content will be transmitted to a third-party endpoint. In this skill's context, automatic daily pushes increase the likelihood of routine exfiltration or accidental disclosure if the webhook target, token, or destination is misconfigured or untrusted.

Session Persistence

Medium
Category
Rogue Agent
Content
```bash
# 编辑 crontab
crontab -e
# 添加一行（修改路径和时间为你配置的）
30 8 * * * cd /path/to/daily-news-push/scripts && python generate_and_send.py
```
Confidence
87% confidence
Finding
crontab -e

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal