Pastewatch MCP
Security checks across malware telemetry and agentic risk
Overview
This skill is a clearly disclosed local secret-redaction integration, though it can inspect sensitive files and optionally run a persistent LLM API proxy if the user enables it.
Install only if you trust the Pastewatch upstream project and want a local secret-redaction layer. Review the external binary/source before use, limit scans to intended directories, protect vault keys and audit logs, and enable the persistent proxy only when you want continuous inspection of outbound LLM traffic.
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
62/62 vendors flagged this skill as clean.