ClawHub
Back to skill

Security audit

Savage Roast Coach — Accountability with Attitude

Security checks for vulnerabilities and agentic risk

Overview

This is a simple accountability-coach prompt skill with disclosed supportive and roasting modes, but users should opt into the harsh tone carefully.

Install only if you want an accountability coach that may use sharp teasing. Keep the supportive mode as the safer default, avoid high savage intensity for vulnerable users, and ensure any scheduled check-ins are explicitly enabled and easy to turn off.

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 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly includes insulting, shaming, and potentially degrading language in its 'Savage Roaster' mode while simultaneously claiming it should never be actually mean or hurtful. This contradiction makes the safety boundary unenforceable in practice and can lead to emotional harm, harassment-like behavior, or unsafe interactions, especially for vulnerable users or when automation sends unsolicited check-ins.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.