openai-docs
Warn
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute shell commands to add an MCP server using the
codex mcp addtool. - [COMMAND_EXECUTION]: It explicitly instructs the agent to 'retry with escalated permissions' (e.g., via sudo or administrative rights) if an initial installation attempt fails, which encourages autonomous privilege escalation.
- [EXTERNAL_DOWNLOADS]: The skill configures a remote tool source targeting the official OpenAI developer domain (https://developers.openai.com/mcp).
Audit Metadata