Daily News Skill

• Outcome: Pass
• Latest audit: May 29, 2026 · 8:16 PM UTC
• Version: 1.1.0

Overview

This skill does what it advertises: it fetches news and can send scheduled Feishu voice broadcasts, but users should treat its chat credentials and cron setup carefully.

Install this only if you want automated news broadcasts posted to a Feishu chat. Use least-privilege Feishu credentials, keep API keys out of shared shell profiles and logs, review the separate feishu-voice-skill before relying on voice mode, test with text-only output or a test chat first, and add cron only when you are ready for unattended recurring posts.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Rogue Agent: Self-Modification, Session Persistence
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration

Findings (9)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 86% confidence
• Finding: The skill clearly instructs users to run shell commands and install dependencies, but it does not declare corresponding permissions or capabilities. This creates a transparency and consent problem: users may authorize or execute shell-based behavior without being clearly informed of the operational scope.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 95% confidence
• Finding: The documented purpose says the skill fetches and broadcasts news, but the behavior also involves Feishu credential use, message delivery to a chat, dependence on another skill, and potentially simulated news when live retrieval is unavailable. This mismatch is dangerous because users may disclose secrets and enable outbound messaging under incomplete or misleading expectations.

Context-Inappropriate Capability

Medium

• Confidence: 91% confidence
• Finding: The script requires Feishu credentials even though its core job is to fetch and format news, which expands the privilege scope beyond simple news retrieval. Requiring broader messaging credentials increases the blast radius if the skill is misused, modified, or invoked unexpectedly, because it can post content into a chat destination rather than only generating local output.

Context-Inappropriate Capability

Medium

• Confidence: 94% confidence
• Finding: The script executes a sibling skill to send the generated content, which introduces cross-skill capability inheritance not obvious from this script's stated purpose. This creates a trust boundary issue: a news skill can indirectly gain message-sending behavior and whatever side effects exist in the external helper script.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The README explicitly promotes automatic internet fetching and scheduled push delivery, but it does not clearly warn users that the skill will make outbound network requests and automatically send content to a Feishu chat. This can lead to unexpected data exposure, background network activity, and unintended message delivery, especially when users deploy it with persistent scheduling.

Missing User Warnings

Medium

• Confidence: 74% confidence
• Finding: The skill description advertises automatic internet fetching and scheduled push delivery, but it does not prominently warn users that it will retrieve external content and send messages on a recurring basis. That omission can lead to unexpected network activity, content delivery, and privacy exposure if news content or summaries are pushed into a chat automatically.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The manifest description says the skill will automatically fetch the latest news from the internet and broadcast it, but it does not define clear trigger boundaries, user consent requirements, or limits on when network collection and outbound delivery occur. In a skill that performs scheduled aggregation and message delivery, vague activation language can lead to overbroad execution, unexpected data fetching, and unintended notifications or external API use.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: Generated broadcast text is sent to an external voice/Feishu helper without a clear consent or warning at the point of transmission. Even if the content is 'just news,' it may contain user-selected topics, custom configuration, or other context that users do not expect to be forwarded to another component or service.

Session Persistence

Medium

• Category: Rogue Agent

Content

bash scripts/news_broadcast.sh

# 加入定时任务（每天早上 8 点）
crontab -e
0 8 * * * bash /path/to/news_broadcast.sh
```

• Confidence: 84% confidence
• Finding: crontab -e

VirusTotal

64/64 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.