ClawHub
Back to skill

Security audit

story-long-scan

Security checks across malware telemetry and agentic risk

Overview

This skill appears to collect public web-novel ranking data and write local analysis reports as described.

Install only if you are comfortable with the agent running Node scraper scripts, opening browser/CDP sessions, visiting public novel-platform pages, and writing Markdown reports. Use an explicit output directory, avoid using a sensitive logged-in browser profile if you do not want site sessions involved, and consider target-site rules before scraping.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill instructs the agent to use network access, browser/CDP automation, and environment-backed script execution, but does not declare those permissions. That creates a transparency and consent gap: a user may invoke what appears to be a simple analysis skill while it actually performs live web scraping and local file writes, increasing the risk of unintended external access and side effects.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
The documented purpose says the skill analyzes ranking data, but the body authorizes broader behavior: scraping additional platforms not named in the description, automating Chrome/CDP, traversing detail pages, handling anti-bot flows, and writing structured output files. This mismatch is dangerous because it hides the real operational scope, making overcollection, unexpected browsing, and unintended filesystem changes more likely under an apparently benign trigger.

Vague Triggers

Medium
Confidence
85% confidence
Finding
The trigger phrases include broad natural-language activators such as generic questions about what is popular, which may match ordinary conversation and invoke network scraping workflows unintentionally. In this skill's context, accidental activation is more dangerous because invocation can lead to automated browsing, cross-site scraping, and local file creation rather than a read-only response.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical
Code
suspicious.dangerous_exec
Location
scripts/cdp-utils.js:88