ClawHub
Back to skill

Security audit

南京留学移民审计 | Nanjing Agency Audit

Security checks for vulnerabilities and agentic risk

Overview

This skill is a purpose-aligned audit guide with an optional local JSON validator and shows no hidden persistence, credential use, or data exfiltration.

Installers should understand that this skill guides an agent to perform public-source due diligence and may process sensitive materials they provide, such as contracts, budgets, screenshots, and candidate agency details. Use care with personal documents, verify mutable legal and registry claims from official sources, and treat generated contract clauses as drafting support rather than legal advice.

SkillSpector

Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Lp3

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding
Without declared permissions the skill's intent is opaque and cannot be validated.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
This is a mismatch because the declared purpose describes a broad substantive auditing capability for Nanjing overseas-agency due diligence, including roster generation, qualification/entity verification, comparisons, investigations, and contract-clause drafting. The actual code only validates the structure and internal consistency of an already-produced JSON audit matrix. It does not access external data sources, perform real-world checks, generate comparisons, investigate complaints, or draft clauses. While the validator supports part of an audit workflow, its primary purpose is much narrower than the declared description.

Natural-Language Policy Violations

Medium
Confidence
92% confidence
Finding
The helper docstring explicitly states 'Return a Chinese-first bilingual message,' and the script consistently uses that function for user-facing output. This imposes a language/locale preference on all users without offering a choice or opt-in, which matches the policy category for forced language behavior.

Static analysis

No suspicious patterns detected.