ClawHub
Back to skill

Security audit

AI KEY·大客户会不会跑

Security checks for vulnerabilities and agentic risk

Overview

The skill performs its advertised customer-risk analysis and saves relevant business lessons, with privacy settings that users should check.

Review the missing configuration and interaction guide before supplying customer or financial details. Confirm the memory location, access permissions, and deletion arrangements. Use customer aliases and explicitly request no saved notes if you want session-only analysis; the bundled instructions specify no retention limit.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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 (1)

The skill persists sensitive business observations without a user-facing privacy or retention warning.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

The skill instructs the agent to persist customer-loss observations, business-specific concentration patterns, and user decisions in reusable memory without an explicit notice, consent step, or retention limit. These records can expose commercially sensitive information if the configured memory is shared, accessed by unrelated agents, or retained beyond the user's expectations. Persistence serves a legitimate analytical purpose, and no external transmission is specified, but report-level customer anonymization does not explicitly protect the stored memories.

Content

No source excerpt is available for this finding.