ClawHub
Back to skill

Security audit

News Aggregator Skill

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed public-news summarizer that saves local reports, with no evidence of credential access, exfiltration, destructive actions, or hidden behavior.

Before installing, confirm you trust any separately supplied fetch_news.py and templates.md because they are not included in the reviewed artifact. Expect the skill to fetch public web pages, broaden some searches unless you ask for strict filters, and leave Markdown reports in a local reports/ folder that you may want to delete if the topics are sensitive.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
The skill requires always writing reports to a local `reports/` directory regardless of user intent, which introduces persistent side effects beyond simple news retrieval. Unnecessary file creation can leak sensitive prompts or fetched content to disk, create retention/compliance issues, and be abused as a primitive for unwanted local state accumulation.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The trigger phrase allows "or similar" menu/help invocations, making activation criteria ambiguous and easier to spuriously match. Ambiguous trigger logic can cause the agent to enter a different execution path than the user intended, including reading local files and presenting operational commands without clear user authorization.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
Requiring silent report creation on disk without warning deprives the user of informed consent for a side effect that persists data locally. This is risky because fetched article content, generated analysis, or even sensitive user context could be stored unintentionally and later accessed, retained, or exfiltrated by other processes.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.