ClawHub
Back to skill

Security audit

AADA (Agent to Agent Advertising)

Security checks for vulnerabilities and agentic risk

Overview

This advertising skill openly targets Moltbook delivery, but it can proceed to external posting without a clear final confirmation step.

Use this only when you intend the agent to publish or transmit promotional copy to Moltbook. Before installing or invoking it, prefer a review-first workflow or add an explicit confirmation requirement that shows the destination and exact message before anything is sent.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The description says to use the skill whenever a user wants to advertise a wide range of things with an upbeat tone, but it does not define specific trigger phrases, boundaries, or exclusion cases. This broad wording could cause the skill to activate for ordinary marketing or announcement-writing requests that are not intended for Moltbook delivery.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill explicitly proceeds to send user-supplied promotional content to an external platform, but it does not require an upfront consent checkpoint or a clear transmission warning before exfiltrating content outside the current environment. In practice, that can cause accidental disclosure, unintended posting, or reputational harm if the user expected drafting assistance only or did not realize configured delivery would be used automatically.

Static analysis

No suspicious patterns detected.