Security audit
Willow External Guard
Security checks for vulnerabilities and agentic risk
Overview
This skill is a focused prompt-injection guard with a small local Python scanner and no evidence of hidden network, credential, or destructive behavior.
Install only if you want Willow-style prompt-injection screening for external content. Treat the scanner as a guardrail, not a complete security boundary: check scan results and exit codes before passing content onward, and do not pipe --wrap output directly into an LLM or KB ingest path unless your caller separately blocks BLOCKED results. Also decide whether local guard-event metadata logging is acceptable in your workspace.
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.