ClawHub
Back to skill

Security audit

图表工坊专业版

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local chart-generation guide with some broad and inconsistent documentation, but no artifact-backed malicious behavior.

Install only if you are comfortable with a skill that can run Python commands and create chart/report files. Clarify whether any networked or API-backed features actually exist, and use explicit output paths to avoid unexpected files in shared or important directories.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The skill documentation makes contradictory trust claims: it says the tool is fully local and requires no network or API keys, yet elsewhere discusses API authentication, rate limiting, timeouts, and network failures. This can mislead operators into granting unsafe trust assumptions, weakening review of outbound communication, credential handling, or data exposure paths.

Vague Triggers

Medium
Confidence
79% confidence
Finding
The invocation guidance is overly broad, encouraging use across many data-analysis and reporting scenarios without narrow trigger conditions. In an agent environment with exec/write tools, vague activation boundaries increase the chance the skill is invoked unnecessarily and performs file-generating or command-executing actions on loosely related prompts.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The markdown promotes automated file generation, export, naming, and archiving behavior without a prominent warning that the skill will create or modify filesystem contents. In a skill that has write and exec capabilities, users may invoke it without understanding that it can populate directories, overwrite outputs, or persist artifacts automatically.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The usage examples include initialization, output directory creation, and export commands that write files, but they lack clear user-facing warnings about file creation, overwrite behavior, and destination paths. Because the skill also exposes exec and write tools, these examples normalize state-changing operations without sufficient consent boundaries.

Static analysis

No suspicious patterns detected.