ClawHub
Back to skill

Security audit

Caption Writer

Security checks for vulnerabilities and agentic risk

Overview

This is a straightforward social caption-writing skill with disclosed use of brand voice files and no executable or persistence behavior.

Before installing, understand that this skill will look for brand-profile.md and voice.md so captions match your brand. That is appropriate for this use case, but avoid using it in workspaces where those files contain sensitive details you do not want used for marketing copy.

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 (1)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill’s trigger phrases are broad enough to match common, everyday writing requests such as 'write a caption' or 'what should I write for this post,' which can cause the agent to invoke this skill in situations where a more specific or safer skill would be more appropriate. Over-broad routing increases the chance of unintended tool use, misclassification of user intent, and accidental access to brand-profile or voice data when the user did not explicitly request brand-specific caption generation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.