Back to skill

Security audit

Meta Ads Lead Gen Analysis

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed, read-oriented Meta ads lead-generation analysis helper with only session-local handoff data for downstream recommendations.

Install only if you are comfortable providing a read-only Meta ads token and ad account ID. The skill may place campaign diagnostics in the current session for a companion recommendation skill, but the artifact does not show account changes, deletion, public posting, hidden persistence, or credential exfiltration behavior.

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
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 (2)

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The manifest frames this skill as specialized analysis for Meta lead generation campaigns, which implies diagnostic/read-oriented behavior. However, the file explicitly says the skill writes multiple keys into session context so another skill can consume them, which is state mutation beyond pure analysis output.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

Line L141 says 'This is analysis only,' which suggests no side effects beyond producing analysis. But lines L115-L126 explicitly define session-context writes, so the documentation contradicts itself about whether the skill has stateful side effects.

Content

No source excerpt is available for this finding.