Back to skill

Security audit

JSON转YAML工具

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple JSON-to-YAML converter but asks for broader powers like command execution, file writing, API keys, and external integrations without clear limits.

Review this skill carefully before installing. It may be usable only if you intend to give it access to files, shell execution, and API credentials, but those powers are not well scoped for a JSON-to-YAML converter. Prefer a converter that performs local data transformation without API keys or command execution.

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (5)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill claims that risky code and external dependencies were removed, yet it still advertises exec, write, API-key, and network-oriented behavior. This mismatch can mislead users and reviewers into granting a broader trust level than warranted, increasing the chance that powerful capabilities are enabled without proper scrutiny.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

A skill framed as a simple JSON-to-YAML converter expands into generic automation, file handling, API integration, and command execution. That scope inflation violates least privilege and can cause an agent or user to invoke a far more capable skill than expected, opening paths to filesystem changes, command abuse, or unintended external communications.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

Command execution is not justified for a JSON-to-YAML conversion task and creates a direct path to arbitrary shell actions if the surrounding agent passes user-controlled or model-generated input into exec flows. In the context of a deceptively simple conversion skill, this hidden power materially increases the risk of abuse and unexpected system impact.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

Requiring external API integration and an API key for a local JSON-to-YAML transformation is unnecessary on its face and expands the attack surface to include credential handling, data exfiltration, and third-party transmission. Because the skill's stated purpose does not require network access, this makes the capability set more dangerous than the user would reasonably expect.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The description uses broad, generic language about API integration, webhooks, system connections, and automation without defining clear invocation boundaries. Ambiguous scope increases the chance that an agent will over-apply the skill in contexts beyond safe JSON-to-YAML conversion, especially given the declared powerful tools.

Content

No source excerpt is available for this finding.