plugin-security-scan

What problem does it solve?

It helps prevent unsafe or malicious behavior in plugin/skill content by detecting prompt injection, credential exfiltration, and privilege-escalation instructions that simple regex checks can miss.

Core Features & Use Cases

• Semantic threat detection: Identifies instruction-level prompt injection patterns, including system-prompt override attempts and persona replacement.
• Credential exfiltration risk spotting: Flags directives to read or transmit secrets via commands or file path targeting (e.g., .env, .ssh).
• Privilege escalation detection: Detects claims or directives that expand authority beyond the intended sandbox/scope.

Quick Start

Ask your EvoNexus plugin install gate to run the plugin-security-scan on the provided plugin markdown files to receive a strict JSON verdict and findings.