Skills
skills/metyatech/skill-post-deploy/post-deploy/Gen Agent Trust Hub

post-deploy

Fail

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The AGENTS.md file contains explicit instructions to bypass platform security restrictions on PowerShell commands (specifically Remove-Item) by using alternative .NET methods like [IO.File]::Delete.
  • [COMMAND_EXECUTION]: The global rules in AGENTS.md direct the agent to use sudo directly for tasks requiring elevated privileges, which is a high-risk instruction that could lead to unintended system-wide modifications.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of several global npm packages from the author's namespace, including compose-agentsmd, @metyatech/task-tracker, and @metyatech/thread-inbox.
  • [PROMPT_INJECTION]: The procedures in SKILL.md involve reading service log files, which creates an indirect prompt injection surface. Since the agent is also instructed to use high-privilege commands like sudo, malicious content in logs could potentially be used to exploit the system.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Apr 9, 2026, 07:24 AM
Security Audit — agent-trust-hub — post-deploy