ClawHub
Back to skill

Security audit

Moltext

Security checks across malware telemetry and agentic risk

Overview

Moltext is a disclosed documentation-ingestion helper; its npm install, URL fetching, local output files, and optional LLM credentials fit that purpose, with normal caution for untrusted docs.

Install only if you are comfortable globally installing the moltext npm package and letting it fetch documentation URLs you choose. Prefer raw mode when no LLM is needed, use scoped API keys if enabling OpenAI, avoid sensitive private sources unless you understand the data flow, and review compiled output before treating it as agent context.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The skill encourages broadly compiling documentation from arbitrary internet URLs into agent memory context without clear trigger constraints, trust boundaries, or source validation. In an agent setting, this can lead to ingestion of untrusted or malicious documentation content that influences downstream behavior, especially because the skill frames the output as authoritative memory to be trusted.

VirusTotal

50/50 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.