baoyu-image-gen
Pass
Audited by Gen Agent Trust Hub on Apr 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Executes a local TypeScript entry point (
main.ts) using thebunruntime to orchestrate API interactions and file management. - [EXTERNAL_DOWNLOADS]: Retrieves generated image data from OpenAI's and Google's official API endpoints.
- [DATA_EXFILTRATION]: Transmits user-provided prompts and reference images to external AI service providers (Google and OpenAI) as part of the intended functionality.
- [PROMPT_INJECTION]: The skill processes content from external files via the
--promptfilesargument, establishing a surface for indirect prompt injection. - Ingestion points:
scripts/main.tsreads text from file paths provided in the--promptfilesCLI argument. - Boundary markers: Absent; contents of multiple files are concatenated and passed directly to the AI provider.
- Capability inventory: The skill has the capability to write to the filesystem (saving images) and perform network operations (API calls).
- Sanitization: No validation or sanitization of the file content is performed before transmission to the model.
Audit Metadata