agent-plugins-skill
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a documentation and reference package designed to assist agents in creating or reviewing plugins following the Agent Plugins 1.0.0 specification. It does not ship with any executables or scripts that run on the host system.
- [SAFE]: The documentation includes explicit security best practices, warning against hardcoding credentials in
plugin.jsonormcp.json(categories: [CREDENTIALS_UNSAFE]) and providing specific logic to prevent path traversal when loading local plugins. - [SAFE]: All provided examples use non-functional, descriptive placeholders and demonstration domains (e.g.,
mcp.example.org,org.example.editor) to illustrate the required JSON structures without initiating actual network connections. - [SAFE]: No obfuscation, privilege escalation, or exfiltration patterns were identified across the 15 files analyzed. The skill effectively acts as a safety-conscious instructional manual.
Audit Metadata