ClawHub
Back to skill

Security audit

AIsa Youtube Search

Security checks for vulnerabilities and agentic risk

Overview

The reviewed artifacts are coherent ClawHub and Convex workflow helpers that disclose their higher-impact commands and do not show hidden exfiltration, persistence, or destructive behavior outside their stated purposes.

Install these skills only for agents you expect to perform ClawHub/Convex maintainer work. Review sensitive commands before use, keep GitHub/ClawHub/Convex tokens least-privileged, use dry-run modes where available, and consider the autoreview --no-yolo option if nested full-access review is not appropriate.

SkillSpector

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

Static analysis

No suspicious patterns detected.