OpenClaw Model Card

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:14 PM UTC
• Version:
  0.1.0

Overview

This skill coherently reads an OpenClaw config to list models and optionally create a local image card, with no evidence of hidden exfiltration or destructive behavior.

Install this if you want a local OpenClaw model inventory or shareable model-card image. Use an explicit --config path when possible, review generated text/images before sharing because they can expose model and provider setup, and only render configs you trust because config-derived values are converted into HTML for the image step.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 91% confidence
• Finding: The skill instructs the agent to execute local Python scripts, read a supplied configuration file, and optionally write an image artifact, which are code-capable behaviors despite no declared permissions. This mismatch is dangerous because it weakens policy enforcement and user visibility: a caller may invoke shell, file read, and file write operations that are not explicitly disclosed or constrained by the skill metadata.

VirusTotal

64/64 vendors flagged this skill as clean.