ClawHub
Back to skill

Security audit

Dental Office Marketing Copy

Security checks for vulnerabilities and agentic risk

Overview

The advertised compliance checker incorrectly clears risky dental claims and patient details, creating misleading reassurance before publication.

Do not treat 'No flags' as evidence that copy is compliant or safe to publish. The checker needs correction and regression testing; retain human review of claim evidence, testimonials, consent, and patient information. Use de-identified drafts. The optional Windows application requires a separate trust decision.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

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 (3)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The reported literal-substring handling of regex patterns would create a fail-open compliance check, potentially letting unsupported dental health claims pass and encouraging misplaced confidence before publication. The script was not supplied, so that implementation defect cannot be independently confirmed; the absence of code that generates copy is not itself a mismatch because the agent can perform that work. Required clinician sign-off and the prohibition on automatic publishing reduce, but do not eliminate, the risk.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_write.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The advertised compliance check reports ordinary risky dental claims as having no flags because it treats detection expressions as literal text.

Medium
Category
Excessive Agency
Confidence
100% confidence
Finding

The rules contain regular expressions, but the checker uses literal substring matching instead of evaluating them, so ordinary claims such as 'painless' or 'cures gum disease' produce a misleading 'No flags' report. This undermines the skill's advertised compliance safeguard and could lead staff to publish unsupported medical claims or sensitive patient details; the dental marketing context increases the consequences, although the script does not itself publish content.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.