ClawHub
Back to skill

Security audit

Install Stack Flagos

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed GPU-container installer, but it will materially change the target container and should be used only where that is acceptable.

Install this only in a disposable or dedicated GPU container, not a shared production environment. Expect it to replace triton, clone and build remote code, install packages in editable mode, and persist FLAGCX_PATH in the container shell startup file. Review the package sources and mirrors first if supply-chain integrity matters, and keep a container snapshot or rebuild path available.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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 (2)

subprocess module call

Medium
Category
Dangerous Code Execution
Content
]
    for cmd, vendor in checks:
        try:
            r = subprocess.run(cmd, shell=True, capture_output=True, timeout=5)
            if r.returncode == 0 and r.stdout.strip():
                return vendor
        except Exception:
Confidence
92% confidence
Finding
r = subprocess.run(cmd, shell=True, capture_output=True, timeout=5)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill performs invasive actions in a live container—package installation, source builds, uninstalls, git clones, and shell profile modification—without a prominent upfront warning that it will modify the target environment. In security-sensitive or production containers, this can lead to unintended state changes, breakage, or persistence of environment changes that the user did not clearly consent to.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.