ClawHub
Back to skill

Security audit

fal.ai

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed fal.ai integration through Maton; it uses network API access and stored authentication, but does not show hidden or destructive behavior.

Before installing, understand that this skill routes fal.ai requests through Maton and may use your fal.ai/Maton account for model runs that can consume credits. Prefer OAuth, confirm any new connection or POST/PUT/PATCH/DELETE request, and avoid running the Python workflow automatically unless you have reviewed the exact request payload and target model.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The skill explicitly states that local execution is out of scope, yet it embeds a Python workflow that runs locally and invokes subprocesses. In an agent setting, such examples can normalize or induce unsafe behavior by encouraging script execution and polling logic outside the constrained API-call model, increasing the chance that external data or future modifications get wired into executable local code.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.