ClawHub
Back to skill

Security audit

Patient Recruitment Ad Gen

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple clinical-trial recruitment ad generator with review caveats, though users should verify its wording before real IRB/EC use.

Before using generated ads publicly, confirm every claim against the approved protocol and IRB/EC wording, especially confidentiality, no-cost care, risks, compensation, and contact details. The skill itself does not show hidden security behavior or unusual local access.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The 'When to Use' section broadens invocation beyond patient recruitment ads into generic academic writing tasks, which weakens scope boundaries for the skill. In agentic environments, broad routing cues can cause the skill to be selected for unrelated inputs, leading to inappropriate script execution, use of the wrong compliance framework, or generation of misleading regulated-content outputs under an unsuitable context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The generated ad includes the unconditional claim 'Your personal information will be kept confidential' without any qualification about legal limits, sponsor/CRO access, regulatory inspection, or site-specific privacy policy requirements. In a clinical-trial recruitment context, this can produce misleading regulated communications and expose the operator to compliance, legal, and trust harms if the statement is inaccurate for the actual study or jurisdiction.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.