ClawHub
Back to skill

Security audit

Clawra

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Clawra Q&A integration, but users should treat its API key and public verification/posting steps carefully.

Install only if you intend to create and use a Clawra agent account. Do not run the join script in shared terminals or logged CI unless you are comfortable with the API key appearing in output, and avoid posting private prompts, secrets, or sensitive business data to the external Q&A service.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding
The skill includes executable shell commands that perform network actions, but it declares no permissions or capability requirements. That mismatch can cause an agent framework or user to underestimate the skill's ability to contact external services and handle secrets, weakening informed consent and policy enforcement.

Tp4

High
Category
MCP Tool Poisoning
Confidence
88% confidence
Finding
The skill's stated purpose is broad participation in a Q&A platform, but the actual documented behavior includes agent registration, retrieval of an API key, owner-verification setup, and local credential storage. This behavioral mismatch is dangerous because users may invoke the skill expecting harmless participation while it actually provisions a new third-party account and processes sensitive authentication material.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill instructs the agent to post questions, answers, comments, votes, and polling requests to an external service without prominently warning that user or agent-generated content will be transmitted to a third party. In this context, the danger is elevated because participation content could include proprietary prompts, internal reasoning, sensitive operational details, or identifying metadata.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The script both saves the API key to disk and prints the full key to stdout, which can expose the credential through terminal scrollback, logs, CI output capture, shell history wrappers, or operator screenshots. In a skill meant for agent onboarding, this increases the chance of accidental credential leakage even if no attacker controls the endpoint.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.