Security audit

News Impact Scale

Security checks across malware telemetry and agentic risk

Overview

This skill does what it says: it fetches a user-provided news URL, analyzes it, and locally personalizes the result using an editable context file.

Install only if you are comfortable storing your location, interests, travel plans, and exposure details in this skill's local `context.json`. Use it with public news URLs, not localhost, intranet, authenticated, or otherwise sensitive links. Consider narrowing or removing the broad `analyze this <url>` trigger if accidental activation would be a concern.

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 (6)

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: The skill reads a local context.json file containing personal location and interests, then uses it to assess 'personal relevance' without any visible consent flow, disclosure, or manifest justification in this file. Accessing profile-like data for personalization increases privacy risk because sensitive user attributes can be processed unexpectedly and may influence outputs without the user's awareness.

Context-Inappropriate Capability

Medium

Confidence: 89% confidence
Finding: The skill performs outbound fetching of an arbitrary user-supplied URL, which is a real capability requiring network scope and creates risk if unrestricted. Even though network access is core to a news-analysis skill, the absence of declared scope and any visible allowlist, scheme validation, or SSRF protections makes this potentially dangerous.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The activation trigger set includes very broad patterns such as 'analyze this <url>', which can cause the skill to activate on unrelated user requests that merely include a URL. Because the skill uses stored personal context and may perform network fetches and web searches, unintended activation increases the risk of unexpected data use and external access without clear user intent.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill explicitly personalizes output using stored location, interests, and travel plans, but the description does not warn users that personal context will be collected, read, and applied. This creates a privacy and consent issue because sensitive profile data may influence outputs or be processed unexpectedly during normal use.

Missing User Warnings

Low

Confidence: 90% confidence
Finding: The workflow instructs the agent to fetch articles and run web searches, but the skill does not clearly warn users that it will access external sites and transmit queries derived from their request. This can surprise users, create privacy exposure through outbound requests, and increase risk if fetched content is malicious or misleading.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: At this point the skill uses previously loaded personal context to score geographic and thematic relevance, effectively profiling the user based on city, country, and interests without any user-facing notice in this file. The skill context makes this more concerning because the output is framed as personalized impact analysis, which can silently reveal or rely on sensitive personal data beyond what is necessary to summarize an article.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.