ClawHub
Back to skill

Security audit

Ai Workflow Red Team Lite

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed defensive AI-workflow review helper that runs a local Python report generator on user-selected input and does not show hidden network, credential, or destructive behavior.

Reasonable to install for defensive AI-workflow review. Use it only with files or text you intend to review, avoid feeding unnecessary secrets or personal data, and choose --dry-run/stdout or a safe --output path when running the Python helper.

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 (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding
The skill declares no explicit permissions, yet its content indicates capabilities to read local resources, write output files, and invoke `python3`, which creates a gap between the trust signal presented to users and the actual power the skill may exercise. Even though the stated use is defensive red-teaming, undeclared file and shell capability increases the chance of unintended local data exposure or execution in environments that rely on manifest permissions for enforcement or user consent.

Tp4

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding
The skill is presented as a narrowly scoped AI workflow red-team assistant, but the observed behavior includes generic local content review features such as arbitrary directory inspection, CSV/TSV summarization, regex-based file scanning, and integrity checks. This broader-than-declared functionality materially increases risk because it can be repurposed to enumerate sensitive local files and extract structured information unrelated to the claimed task, making users and reviewers less likely to apply appropriate scrutiny.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.