ClawHub
Back to skill

Security audit

improve-codebase-architecture

Security checks across malware telemetry and agentic risk

Overview

This skill performs a disclosed architecture-review workflow and only has limited, purpose-aligned documentation side effects.

Install only if you are comfortable with an architecture-review skill reading relevant project files, using sub-agents, and making limited CONTEXT.md glossary edits during the conversation. Review those documentation changes like any other repo change.

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
91% confidence
Finding
The skill explicitly authorizes modifying `CONTEXT.md` inline during the conversation without first requiring an explicit user confirmation at the moment of the write. Even though the edits are framed as glossary updates rather than production code, this still permits persistent repository changes from a review-oriented skill, creating a risk of unexpected file modification, prompt-injected documentation changes, or drift in project records without clear user consent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.