ClawHub
Back to skill

Security audit

Internalize Me

Security checks across malware telemetry and agentic risk

Overview

This appears to be a documentation-focused skill whose file changes are aligned with building and reviewing repository docs.

Install this if you want an agent to help review or improve technical documentation. Before using apply-fixes mode on an important repository, make sure you are comfortable with it editing docs, AGENTS/CONTRIBUTING files, README files, and possibly compatibility symlinks; use report-only mode when you only want findings.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Description-Behavior Mismatch

Medium
Confidence
92% confidence
Finding
The skill explicitly instructs the agent to create and update files inside the analyzed repository under `.learn/<project-slug>/`. That exceeds a read-only analysis scope and can cause unintended workspace modification, especially when the user asked only for analysis and was not clearly warned that files would be written.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill directs output into the target project's `.learn/` directory without an explicit user-facing warning at execution time that local files will be modified. This creates a consent and transparency problem, and may surprise users by changing repositories, dirtying git state, or overwriting prior notes.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.