Security audit
Semantic Circuit Breaker
Security checks across malware telemetry and agentic risk
Overview
This is a markdown-only guidance skill about semantic circuit breaker design, with no executable behavior or sensitive access.
This appears safe to install as a guidance-only skill. If a future version adds real traffic interception, response logging, embedding-provider calls, or fact-checking integrations, review those additions for privacy, retention, and operational controls.
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
VirusTotal
64/64 vendors flagged this skill as clean.
Static analysis
No suspicious patterns detected.