ClawHub
Back to skill

Security audit

Self Improving Agent

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed local note-taking aid for agent learnings, with no evidence of hidden network access, credential use, or destructive behavior.

Install this only if you want the agent to keep local project memory about mistakes, corrections, and feature requests. Review `.learnings/` and any promoted memory files periodically, and avoid letting sensitive details, credentials, or private conversation content be logged.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill instructs the agent to create and append to multiple files in the user's workspace during routine use, but it does not require explicit user confirmation or a warning that local project files will be modified. In an agent setting, silent workspace writes can violate user expectations, create persistence artifacts, and capture sensitive conversational or operational context in repository files.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.