Security audit
skill-architecture-standard
Security checks for vulnerabilities and agentic risk
Overview
This is a non-executable reference skill that gives architecture and safety guidance for creating or reviewing OpenClaw skills.
Before installing, understand that this skill may make an agent more structured and stricter when helping create or review skills, but the reviewed artifact does not add executable code, hidden persistence, or credential handling.
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.