ClawHub
Back to skill

Security audit

CSV解析生成-免费版

Security checks for vulnerabilities and agentic risk

Overview

This CSV skill is mostly documentation, but it asks for command execution and describes API/key handling and broad file/API/command workflows that do not fit a simple CSV parser/generator.

Review this skill before installing. It does not show malicious code or persistence, but it grants and discusses capabilities broader than a simple CSV helper. Only use it in an environment where command execution is controlled, and do not provide API keys unless a specific, trusted CSV-related workflow actually requires them.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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
Findings (4)

Description-Behavior Mismatch

High
Confidence
95% confidence
Finding
The skill is presented as a narrowly scoped CSV parser/generator, but the documentation later advertises generic file handling, API integration, and command execution. This scope expansion is dangerous because it can cause an agent or user to invoke powerful capabilities under the guise of a low-risk data-formatting skill, increasing the chance of unintended file modification, external data exfiltration, or shell misuse.

Context-Inappropriate Capability

Medium
Confidence
89% confidence
Finding
A CSV parsing/generation skill does not need external API access or API key handling, so these instructions introduce unjustified secret management and network-related trust assumptions. Even without explicit exfiltration logic, normalizing API-key setup in an unrelated skill can mislead users into exposing credentials and broaden the operational attack surface.

Context-Inappropriate Capability

High
Confidence
94% confidence
Finding
Documenting command execution as a capability for a simple CSV skill is an unnecessary privilege escalation. In agent environments, exec access materially increases risk because user-supplied file names, CSV content, or task parameters may later be incorporated into shell commands, enabling filesystem damage, data exposure, or command injection chains.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The markdown promotes file writing and command execution without clear warnings, constraints, or approval expectations for those side effects. In an agent setting, unclear disclosure of write/exec behavior can lead users to invoke a seemingly harmless CSV skill that performs impactful actions on the local environment.

Static analysis

No suspicious patterns detected.