ClawHub
Back to skill

Security audit

Mindflow

Security checks across malware telemetry and agentic risk

Overview

Mindflow is a coherent local mind-map image generator, with normal cautions around installing JavaScript packages and rendering generated HTML in Puppeteer.

Install only if you are comfortable with Node/Bun dependencies and local Puppeteer rendering. Use the skill on text, Markdown, or TXT files you intend to convert, and avoid running the bundled HTML renderer on arbitrary untrusted HTML unless network access is blocked or the HTML is sanitized.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The script renders attacker-controlled HTML with Puppeteer via `page.setContent(...)` in a full browser context and does not block external resource loads. That allows embedded `<img>`, `<script>`, `<link>`, fonts, or fetch/XHR-capable page code to trigger outbound network requests during conversion, which can be abused for SSRF-style access to internal endpoints, unexpected data exfiltration, or privacy-leaking callbacks. In this skill’s context, the stated purpose is converting text/Markdown/TXT into mind-map images, so permitting arbitrary HTML with network-capable browser rendering is not necessary and increases risk.

Vague Triggers

Medium
Confidence
83% confidence
Finding
The trigger condition is extremely broad ('whenever text needs to be converted into image-based mind maps'), which can cause the skill to activate on loosely related requests and process unintended user content or files. In an agentic environment, over-broad invocation increases the chance of unnecessary file handling, command execution, and data exposure through downstream tooling, especially because this skill writes temp/output files and shells out to external tools.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.