ClawHub
Back to skill

Security audit

甲供材料与分包领用核对（免费版）

Security checks for vulnerabilities and agentic risk

Overview

This is a local, disclosed ledger-checking skill with a real limitation around omitted paid-version checks, but the limitation is stated in the instructions and runtime output.

Install only if you need the free local checks listed by the skill. Do not treat a clean result as coverage for withheld checks such as over-issue deductions, return-not-offset detection, contract-price mismatch, or missed deductions; review those separately or use a tool that explicitly performs them.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
The skill’s trigger words and top-level description imply support for scenarios like '超领扣款' and '退库没冲减', but the body later states those checks are out of scope in the free version. In a financial reconciliation workflow, this can mislead users into relying on the tool for controls it does not actually perform, causing missed billing or deduction errors rather than direct system compromise.

Ae1

High
Category
analysis-evasion
Content
node scripts/run.mjs --sample # 先看样例
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
node scripts/run.mjs --sample # 先看样例
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Description-Behavior Mismatch

High
Confidence
98% confidence
Finding
The skill metadata and description advertise checks for '超领扣款' and '退库没冲减', but the implementation explicitly lists those checks as withheld and never executes them. This creates a deceptive security/control gap: users may rely on the tool to detect financially material discrepancies that it does not actually test, causing false assurance and missed overcharges or under-deductions.

Natural-Language Policy Violations

Low
Confidence
90% confidence
Finding
The primary display name, summary, and main descriptions are all in Chinese, which effectively sets a fixed language for the skill’s invocation and presentation. Although an English field exists later, the skill does not state that users can choose their language or that Chinese is a required locale for a region-specific workflow.

Natural-Language Policy Violations

Low
Confidence
95% confidence
Finding
The user-facing description, help text, status messages, and usage instructions are entirely in Chinese, indicating the skill is designed to operate in a single language. Under the policy, forcing a specific language without offering the user a language or locale choice is a natural-language policy concern.

Static analysis

No suspicious patterns detected.