ClawHub
Back to skill

Security audit

4skill

Security checks across malware telemetry and agentic risk

Overview

This PRD skill is mostly documentation, but it includes an unattended coding-agent loop that bypasses permission prompts and can repeatedly modify a repository.

Install only if you want PRD templates plus agent-execution guidance. Do not run the unattended permission-skipping loop in a real project; use a separate branch or worktree, keep approvals enabled, review commits, and avoid placing secrets or untrusted instructions in `progress.txt`.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (2)

Natural-Language Policy Violations

Low
Confidence
87% confidence
Finding
The markdown instructs users to use a branch name prefixed with `ralph/`, which imposes a specific organizational naming convention in natural language. No opt-in, alternative, or justification is provided, so this reads as a forced policy constraint rather than an optional example.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
This markdown file includes a fully unattended execution pattern using `claude --dangerously-skip-permissions`, which can lead to autonomous file edits and command execution. The surrounding text presents it as a usage pattern but does not warn users about the elevated risk to system integrity or the need for careful review before use.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.