imagegen
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided text to construct prompts for the Codex image generation tool, creating a surface where malicious instructions in the user input could potentially influence the tool's behavior.
- Ingestion points: The user's image description provided in the skill argument in SKILL.md.
- Boundary markers: The skill instructions suggest appending operational commands to the prompt but do not define strict delimiters for the user-supplied content.
- Capability inventory: The skill has access to Bash, Read, and Write tools to execute the Codex CLI and manage files.
- Sanitization: The skill extracts the destination path from the input but does not sanitize the core image description text before it is written to a temporary prompt file.
- [DYNAMIC_EXECUTION]: The skill dynamically generates shell commands and scripts at runtime to interact with the Codex CLI.
- The skill constructs command strings incorporating a thread ID retrieved from tool output and absolute paths resolved by the agent.
- It writes temporary prompt files and reads output from JSONL files to manage the image generation lifecycle.
Audit Metadata