ClawHub
Back to skill

Security audit

Geo Ai Plugin Builder

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent strategy and template helper for designing GEO-aware AI plugin specifications, with no evidence of hidden execution, data exfiltration, persistence, or destructive behavior.

Installers should treat this as a planning/template skill. It may encourage broad AI tool descriptions, so review generated plugin schemas for narrow invocation criteria, clear data boundaries, authentication expectations, and privacy constraints before implementing them in a real product.

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
Confidence
94% confidence
Finding
The manifest description instructs invocation for broad concepts like building, designing, packaging, or standardizing AI plugins/tools, which can overlap with many general product, engineering, or content tasks. In an agentic environment, this can cause the skill to be selected outside its intended GEO/plugin-strategy niche, leading to inappropriate prompt injection of its workflow and guidance into unrelated conversations.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The description examples say to use the tool 'whenever the user asks to compare' and 'when the user provides concrete numbers and wants…', which are broad natural-language triggers that could overlap with many ordinary requests. The file does not provide negative examples or scope limits to clarify when these patterns should not activate a tool.

Static analysis

No suspicious patterns detected.