codex-image-gen
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes external command-line tools including
codexandsips. It also runs a local Python script to read~/.codex/sessions/rollout logs to recover generated image data.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external image prompts without sanitization.\n - Ingestion points: Prompts written to
.tmp/img-prompt.txtinSKILL.md.\n - Boundary markers: Absent in shell interpolation.\n
- Capability inventory: Shell command execution and file writing.\n
- Sanitization: Absent.
Audit Metadata