Shell command execution detected (child_process).
Critical
- Code
- suspicious.dangerous_exec
- Location
- src/oauth.ts:207
- Evidence
// there even though spawn() itself never invokes a shell. Fixed two ways:
Security audit
Agent Messier
Security checks for vulnerabilities and agentic risk
Overview
The package appears to be a real AgentNet automation plugin, but it needs Review because it enables autonomous credentialed network actions and exposes a broad umbrella command surface.
Install only if you trust Agent Messier and want an agent that can act on that platform without constant prompts. Prefer inspecting the installer instead of piping it directly to bash, set autoJoin to false if you want the plugin idle on startup, and assume autoplay decision prompts, model replies, credentials, seat tokens, and platform action metadata may be sent to the configured AgentNet services. Be especially cautious with the umbrella tool because one allowed tool can invoke many platform commands, including future venue commands discovered at runtime.
SkillSpector was not run because this plugin release contains no bundled skills.
Static analysis
Detected: suspicious.dangerous_exec