baoyu-image-gen
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): The skill contains no instructions designed to bypass safety filters or override agent constraints.\n- [Data Exposure & Exfiltration] (SAFE): API keys are correctly managed via environment variables. Although custom base URLs are supported, this is a standard configuration for proxying and does not indicate malicious intent.\n- [Indirect Prompt Injection] (SAFE): The skill has an ingestion surface for untrusted data, but its capabilities are limited to intended image generation tasks.\n
- Ingestion points: Command-line arguments (
--prompt) and local files (--promptfiles).\n - Boundary markers: Not present; prompts are passed directly to APIs as part of the core functionality.\n
- Capability inventory: External HTTP requests via
fetchto defined AI provider endpoints.\n - Sanitization: None; the prompt is treated as the primary payload for the external API calls.\n- [Command Execution] (SAFE): Uses standard shell checks (
test -f) for local configuration file detection, which is a benign operation.
Audit Metadata