Security audit

Web Scraper Pro

Security checks for vulnerabilities and agentic risk

Overview

The skill does not show malware-like behavior, but it presents itself as a real-time market and web scraping tool while returning hardcoded sample data.

Review this before installing if you need real market or news data. It appears suitable only as a demo/static-output helper unless the publisher clearly relabels it or implements real retrieval with source attribution and freshness warnings. Do not rely on its prices, fund-flow, sector, or news outputs for financial decisions.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Tp4

High

Category: MCP Tool Poisoning
Confidence: 94% confidence
Finding: The skill description claims comprehensive real-time scraping and market-data retrieval, but the analyzed behavior reportedly uses static or stubbed functionality instead. This mismatch is dangerous because users may make trust, compliance, or operational decisions based on false assumptions about what data is being fetched and how the skill behaves.

Lp3

Medium

Category: MCP Least Privilege
Confidence: 94% confidence
Finding: The skill advertises network-capable functionality and a command-dispatch tool, but it declares no explicit tool scope such as permissions or allowed-tools. This weakens the trust boundary for users and reviewers because the skill can initiate external requests without clearly constraining or documenting that capability.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill exposes a custom URL scraping command but does not warn users that it can perform outbound requests to arbitrary external sites. That omission can hide privacy, SSRF-like, compliance, and data-exfiltration risks, especially in environments where agents may have network reachability to sensitive internal resources.

Natural-Language Policy Violations

Medium

Confidence: 97% confidence
Finding: The file’s user-facing description and command output are primarily in Chinese, and the CLI usage/help strings throughout the skill do not offer any user opt-in for language selection. Under the policy, forcing a specific language without user choice is a natural-language policy violation unless the locale limitation is clearly justified.

Description-Behavior Mismatch

Medium

Confidence: 97% confidence
Finding: The skill is presented as a comprehensive web scraping and market-data tool, but it only returns hardcoded sample data and performs no live retrieval. This is dangerous because users or downstream agents may make decisions based on stale or fabricated-looking output while believing it is current external data, creating integrity and trust risks rather than direct code-execution risk.

Intent-Code Divergence

Low

Confidence: 88% confidence
Finding: The inline comment claims that actual usage would call a browser or API, but no such behavior exists in the implementation. This discrepancy can mislead reviewers, users, or agent orchestrators into overtrusting the freshness and provenance of the data, compounding the deceptive behavior already present in the tool context.

Static analysis

No suspicious patterns detected.