ClawHub
Back to skill

Security audit

Hourly Knowledge

Security checks for vulnerabilities and agentic risk

Overview

This skill matches its stated purpose of sending hourly facts, with manageable caveats around automatic messages, local topic history, and a helper command that can print its account ID config.

Install only if you want automatic hourly fact messages from the configured bot account. Confirm the accountId is appropriate, ensure users have opted in to recurring messages, and avoid using or exposing the helper's config command unless you are comfortable revealing that identifier.

SkillSpector

Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding
The skill instructs reading and updating a local history file, which implies file-write capability, but no corresponding permission is declared. Undeclared state mutation is dangerous because it bypasses least-privilege review and can let a seemingly simple content-generation skill persist or alter local data unexpectedly.

Tp4

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding
The finding indicates behavior beyond the stated purpose, including reading and outputting local configuration, exposing command-style interfaces for history/config access, and returning random preset categories. Hidden access to config data is especially risky because config files commonly contain sensitive identifiers or secrets, and functionality mismatch undermines trust and reviewability.

Static analysis

No suspicious patterns detected.