ClawHub
Back to skill

Security audit

Fluid Network Agent

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local fluid-network calculator whose file access, Python execution, and optional report output match its stated purpose.

Install only if you want a Chinese-documented local fluid-network analysis helper. Review any TOML the agent drafts or edits, and be deliberate with the optional --output path because the script can write a report there.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (6)

Tp4

High
Category
MCP Tool Poisoning
Confidence
87% confidence
Finding
The core declared computational behavior mostly aligns: the code does solve steady-state network pressures, flows, pressure drops, and optional velocities for scenario-dependent conditions. However, the description also claims TOML input handling and reporting of connectivity and load reliability (PASS/FAIL), which are not implemented in this chunk. Conversely, the code contains additional materially relevant behavior not mentioned in the description: support for nonlinear/quadratic resistance models via a Newton solver and scenario-controlled edge gating through effective resistance changes. Because the declared description includes important outputs/capabilities absent from the supplied code and omits notable implemented solver behavior, this is a description-behavior mismatch.

Lp3

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

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The skill's user-facing description is written as an unconditional Chinese instruction and does not indicate that other languages are supported or that the user can choose a preferred language. This is a natural-language policy concern because it imposes a locale/language choice without opt-in.

Natural-Language Policy Violations

Medium
Confidence
90% confidence
Finding
This markdown file presents all user-facing instructions and operational guidance exclusively in Chinese. Under the policy rules, forcing a specific language without offering the user a locale choice or documenting a justified region-specific constraint is a natural-language policy violation.

Missing User Warnings

Low
Confidence
82% confidence
Finding
The skill says the agent should complete a missing TOML into a runnable version, which can affect user data or configuration content, but the markdown does not warn the user that input files/specifications may be created or altered. For markdown-scoped warnings, user-visible disclosure should mention this behavior when it can change project artifacts or structured inputs.

Natural-Language Policy Violations

Low
Confidence
90% confidence
Finding
This code file contains a natural-language comment written only in Chinese, with no indication that the skill is intentionally region-specific or that alternative language support is provided. The policy requires avoiding forced language or locale constraints unless they are documented and justified.

Static analysis

No suspicious patterns detected.