codex-imagegen

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses codex exec to perform tasks, constructing shell commands that interpolate user-controlled file paths and prompt text. It attempts to mitigate injection via single-quoted heredocs and sandboxing.\n- [PROMPT_INJECTION]: The skill processes user-supplied instructions and passes them to a delegated agent (codex), creating a surface for indirect prompt injection. A malicious user could provide input designed to override the behavior of the downstream image generation process.\n- [EXTERNAL_DOWNLOADS]: The skill includes instructions to install an external MCP server (codex mcp-server), which requires the presence of the codex CLI tool. This adds an external dependency to the agent's environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 02:16 PM
Security Audit — agent-trust-hub — codex-imagegen