ClawHub
Back to skill

Security audit

LGD Fin Guard（金融 AI 合规守门）

Security checks for vulnerabilities and agentic risk

Overview

The skill recommends automatic financial gating using a keyword checklist that can approve descriptions of missing safeguards.

Use this only as a preliminary checklist. Its PASS result does not verify licences, evidence, or regulatory compliance, and its documented enforcement modes do not exist. Before connecting its output to financial workflows, require independent evidence checks, explicit approval of affected flows, and a recovery plan. The bundled code itself does not control transactions; any external enforcement remains unspecified.

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 workflow automatically allows or blocks financial data flows without warning users about the operational impact.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

This workflow step instructs automatic enforcement of the assessment result, rather than merely describing an available tool. Although the skill describes gating as its purpose, it does not warn that enforcement can interrupt legitimate financial processing or permit sensitive workflows based on an unvalidated assessment; the security-scan disclaimer does not explain these consequences.

Content

No source excerpt is available for this finding.