ClawHub
Back to skill

Security audit

JSON转YAML工具

Security checks for vulnerabilities and agentic risk

Overview

This JSON-to-YAML skill asks for broader API, file, command-execution, and credential-related authority than its stated conversion purpose explains.

Review this skill before installing. For a JSON-to-YAML converter, prefer a skill that performs local parsing and output only, or install this one only if you intentionally want its broader read/write, command, and API-driven workflow and can constrain what data, files, credentials, and commands it may use.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (5)

Description-Behavior Mismatch

High
Confidence
97% confidence
Finding
The skill is presented as a narrow JSON-to-YAML converter, but its documented capabilities include file handling, API integration, command execution, and information retrieval. This scope expansion is dangerous because it can cause an agent or user to grant or invoke far more powerful behaviors than expected, increasing the chance of data exfiltration, unintended system changes, or misuse under a misleading label.

Description-Behavior Mismatch

High
Confidence
98% confidence
Finding
The operational instructions tell the agent to establish API connections and execute commands even though the stated purpose is simple data format conversion. This mismatch creates a deceptive trust boundary: operators may approve the skill for low-risk conversion work while it guides higher-risk actions that can touch external systems or the local environment.

Context-Inappropriate Capability

High
Confidence
99% confidence
Finding
System command execution is not justified for a JSON-to-YAML conversion skill and materially raises the attack surface. If an agent follows this skill, command execution could be leveraged to run unsafe local actions, access sensitive files, or chain with user-controlled input into command injection scenarios.

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
External API integration is unnecessary for local JSON-to-YAML conversion and introduces avoidable outbound data flow. This can expose sensitive input data to third parties, create dependency on untrusted services, and expand the skill beyond its expected trust model.

Intent-Code Divergence

Medium
Confidence
91% confidence
Finding
The documentation claims that original risky code and dependencies were removed, yet later sections advertise command execution and API-driven behavior that reintroduce the same classes of risk. Such reassuring language can mislead reviewers into lowering scrutiny and approving a skill whose effective capabilities remain dangerous.

Static analysis

No suspicious patterns detected.