ClawHub
Back to skill

Security audit

yansheng-news-sentiment

Security checks across malware telemetry and agentic risk

Overview

This educational finance-news skill contacts a public finance site but produces sentiment numbers using crude fixed math that is not clearly reflected in its normal output.

Install only if you want a classroom/demo heuristic. Do not rely on its scores for trading, investment, compliance, or reputation decisions without replacing the analyzer or independently verifying the underlying news and sentiment logic.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
86% confidence
Finding
The skill documentation advertises use of external finance sources and the invocation implies network access, but no permissions are declared. Undeclared network capability is dangerous because users and hosting platforms cannot accurately assess data egress, third-party contact, or compliance risk before execution.

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding
The documented behavior promises news collection, sentiment scoring, and event-driven analysis, but the finding indicates the implementation instead scrapes external pages and fabricates sentiment outputs using fixed ratios without performing the claimed analysis. This is dangerous because it can mislead users into trusting invented financial-analysis results, creating integrity, decision-making, and compliance risk, especially in an investment-related context.

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
The script advertises news sentiment collection, scoring, and event-driven analysis, but it only counts occurrences of the string 'news' in fetched HTML and derives fixed sentiment ratios from total article count. This can mislead downstream users or agents into treating fabricated analytics as real market intelligence, creating integrity and decision-making risk in a finance-related context.

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The file description says it is a 'news sentiment analysis script,' but the implementation does not inspect article sentiment or event content at all. In a financial analysis skill, this mismatch increases the chance that users or autonomous agents will rely on inaccurate outputs as if they were analytically grounded.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.