ClawHub
Back to skill

Security audit

Take Back My Latte

Security checks for vulnerabilities and agentic risk

Overview

This skill locally analyzes user-provided OpenAI cost/usage JSON and does not show hidden upload, credential use, persistence, or destructive behavior.

Install only if you are comfortable providing local OpenAI Costs and Usage JSON exports to the agent for local analysis. The inspected artifact does not request API keys or upload data, but the exports may contain sensitive spend, project, model, or API-key identifier details, so use files you intend to analyze.

SkillSpector

Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (1)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill instructs the agent to read user-provided local JSON files and also includes an external website link, while the metadata declares no explicit permissions. This mismatch can undermine permission enforcement and user expectations: a host may allow the skill to operate without surfacing that it reads files and may reference networked resources, increasing the risk of unintended data exposure or unsafe execution contexts.

Static analysis

No suspicious patterns detected.