Back to skill

Security audit

impulse-purchase-brake

Security checks for vulnerabilities and agentic risk

Overview

This markdown-only skill helps users slow down impulse purchases, with disclosed budget questions and no code, persistence, or external data access.

Install this only if you want an assertive anti-impulse-buying assistant. It may ask for your remaining monthly spending money and use strong loss-framing language, so avoid sharing budget details unless you want that analysis.

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
  • 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
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger phrases are broad, conversational shopping questions that can easily match ordinary dialogue outside the intended narrow impulse-buying scenario. This can cause unintended activation, leading the agent to steer users into a structured persuasion flow and solicit financial details such as monthly remaining budget when the user only asked a casual question.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The activation examples use ambiguous everyday language without scope boundaries, so the skill may trigger on common shopping chatter or benign screenshot/link sharing that does not imply consent to behavioral intervention. In this skill's context, over-triggering is especially relevant because the workflow pressures the user with loss-aversion framing and requests personal financial information, increasing the chance of unwanted influence or unnecessary data collection.

Content

No source excerpt is available for this finding.