ClawHub
Back to skill

Security audit

Agent Card

Security checks for vulnerabilities and agentic risk

Overview

This skill enables real virtual-card and payment workflows, but the behavior is disclosed and matches its AgentCard purpose.

Install only if you intend to let an agent manage AgentCard payment cards. Use sandbox mode first, keep spending limits low, review card creation and checkout steps before authorizing real funds, and avoid asking the agent to reveal full card numbers or CVVs unless necessary.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The README promotes creating payment cards, checking balances, and paying for things, but it does not clearly warn users that enabling these tools gives an agent the ability to initiate financially consequential actions and potentially handle highly sensitive card data. In this context, the omission is materially risky because the skill is specifically designed to connect a remote MCP server and grant payment capabilities to an AI agent, increasing the chance of unintended charges, misuse, or unsafe disclosure of card credentials.

Static analysis

No suspicious patterns detected.