Back to skill

Security audit

Launch Blitz

Security checks for vulnerabilities and agentic risk

Overview

This skill is coherent but should be reviewed because it tells an agent to auto-submit public launch content across many external platforms without clear final approval controls.

Install only if you are comfortable with an agent preparing public launch materials for many third-party sites. Before running it, explicitly tell the agent whether you want drafts only or live submissions, choose the target platforms, review all generated content, and require confirmation before anything is posted or submitted.

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
  • 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 (3)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill promotes preparing and submitting product information to many third-party platforms but does not warn the user that business details, URLs, pricing, and competitive information may be transmitted externally. This creates a meaningful risk of unintended data disclosure, premature public launch, and irreversible publication across multiple sites.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The trigger phrase "Run the launch blitz" is broad and does not limit the action to drafting content versus actually performing external submissions. In a skill explicitly oriented around submitting to 21 external platforms, an ambiguous invocation can cause the agent to initiate high-impact actions without first confirming scope, permissions, or whether the user wants automation versus planning only.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The checklist specifies an "X launch thread (English)" without clear user choice, which can cause the agent to generate or prioritize a public-language output the user did not request. While lower severity than direct submission issues, it still overrides user preference and may lead to unintended public-facing content in a specific language.

Content

No source excerpt is available for this finding.