ClawHub
Back to skill

Security audit

MindStudio To API Custom Function Skill

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only skill for generating MindStudio API integration snippets, with expected API-key and external-call guidance but no hidden installer, persistence, or deceptive behavior.

Before installing, review generated functions for the exact endpoint, request fields, credential scope, and output variable names. Use least-privilege API keys and confirm inputVariable fields are read from ai.config while only outputVariableName fields are used as ai.vars keys.

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
97% confidence
Finding
The skill gives inconsistent instructions about where workflow-supplied inputs should come from: earlier it says `inputVariable` values are resolved into `ai.config.*`, but later it says values set via workflow variables should be read from `ai.vars`. This ambiguity can cause generated code to source data from the wrong channel, leading to broken integrations, accidental use of stale variables, or unintended data exposure if sensitive workflow state is read instead of the intended explicit input.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
EXAMPLES.md:14

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
SKILL.md:52