ClawHub
Back to skill

Security audit

Security Digital Human Ops

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only digital-human marketing assistant with no executable code, persistence, credential access, or hidden behavior, though its broad trigger keywords may activate it in unrelated content requests.

Before installing, consider narrowing the trigger keywords so the assistant is used for digital-human banking or insurance marketing requests rather than generic live-streaming or content-creation tasks. Content it generates should still be reviewed for financial advertising and compliance requirements.

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
Confidence
95% confidence
Finding
The trigger list is overly broad and includes generic terms such as content creation, short video, live streaming, and metaverse-related phrases that can match many unrelated user requests. This can cause unintended activation in benign conversations, leading the skill to inject financial-marketing-oriented guidance where it was not requested, which is especially sensitive in banking and insurance contexts.

Static analysis

No suspicious patterns detected.