ClawHub
Back to skill

Security audit

Churn Prevention

Security checks for vulnerabilities and agentic risk

Overview

This is a churn-prevention guidance skill with no executable code, hidden persistence, or automatic account changes.

Before installing, treat this as business-process guidance: review any cancellation-flow or billing-provider changes for legal compliance and require explicit user approval before using Stripe, CRM, email, or analytics tools to change real customer accounts.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill’s invocation guidance is very broad and includes many generic phrases like 'retention' and 'customers are leaving,' which can cause the agent to select this skill in conversations that are only loosely related. Over-selection increases the chance the agent applies churn-prescriptive guidance in the wrong context, potentially leading to irrelevant actions, poor advice, or unintended workflow/tool usage.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This JSON manifest-like file includes expected behavior and assertions that the skill 'Should trigger on casual phrasing' and 'Triggers on casual phrasing.' That activation guidance is overly broad and lacks scope limits or negative examples, increasing the risk of unintended invocation from everyday language.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · references/cancel-flow-patterns.md (reported line 310)May include surrounding context.

md
- Inform users about data retention period post-cancel
- Offer data export before account deletion
- Honor deletion requests within 30 days
- Don't use post-cancel data for marketing without consent

### General Best Practices
- Always show a clear path to complete cancellation