ClawHub

PRD to Prototype

Security checks across malware telemetry and agentic risk

Overview

This skill is an instruction-only PRD and prototype generator with disclosed workspace file creation and optional deployment, but no hidden code, credential access, or destructive behavior.

Install this only if you want product ideas to be turned into local PRD and prototype files automatically. Review /workspace/docs/prd.md before confirming prototype generation, and avoid confidential product details unless you are comfortable with a possible deployed preview link.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The skill advertises activation for very common phrases such as wanting to build or design something, which can cause it to trigger during ordinary product discussions where the user did not explicitly request this workflow. Over-broad triggering increases the chance of unintended file creation and prototype-generation behavior, especially because the skill is designed to take action immediately and write artifacts.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation condition covers essentially any product idea, need, pain point, or creative thought, making accidental invocation highly likely. In context, this is more dangerous because the skill also mandates zero-question execution and immediate generation of files, so ambiguous activation directly leads to unintended actions.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill instructs the agent to write to /workspace and deploy prototypes, but it does not require any user-facing warning or consent about those side effects. This is risky because users may think they are only brainstorming, while the agent can create files and potentially publish content externally once the workflow is triggered.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal