ClawHub
Back to skill

Security audit

JSON检查工具

Security checks for vulnerabilities and agentic risk

Overview

This JSON linter is only a Markdown skill, but it asks for broader authority and API-key setup than its local validation purpose explains.

Install only if you are comfortable with a broadly worded JSON-linting skill that may lead the agent to use exec/write permissions and asks for an API key without explaining why. Prefer a local-only JSON linter or revise this skill to remove the API-key, network, write-access, and generic automation language.

SkillSpector

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 (4)

Intent-Code Divergence

Medium
Confidence
87% confidence
Finding
The skill is presented as a simple local JSON syntax validator, but the documentation also introduces API-key handling, server/network behavior, and command-execution concerns that are not necessary for that purpose. This mismatch can mislead users about the actual trust boundary and capabilities of the skill, increasing the risk of unnecessary secret exposure or unsafe execution in environments that expected a harmless linter.

Intent-Code Divergence

Medium
Confidence
84% confidence
Finding
The output section claims the skill only returns a JSON lint report, yet later sections discuss command failures and server/network issues, implying behavior beyond passive local validation. Such contradictory documentation can cause operators to approve or invoke the skill under false assumptions, which is dangerous when tools include exec and potential remote dependency paths.

Context-Inappropriate Capability

Medium
Confidence
90% confidence
Finding
Requiring an API key and referencing network/server dependencies for a JSON syntax checker is unjustified on its face and expands the attack surface without clear need. Users may provide credentials or permit outbound access unnecessarily, enabling data exfiltration, misuse of secrets, or execution in more privileged contexts than the task requires.

Vague Triggers

Medium
Confidence
80% confidence
Finding
The description frames the skill as applicable to broad productivity, automation, and workflow-optimization scenarios rather than narrowly to JSON syntax validation. Overbroad invocation criteria can cause the agent to select a skill with read/exec/write capabilities in contexts unrelated to its stated purpose, increasing the chance of unnecessary file access or command execution.

Static analysis

No suspicious patterns detected.