imagen
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied natural language prompts to drive image generation, creating a potential surface for indirect prompt injection.
- Ingestion points: Described in SKILL.md as taking a text prompt describing the desired image (e.g., 'A futuristic city skyline').
- Boundary markers: There are no visible delimiters or instructions to the agent to treat the prompt content as untrusted data or to ignore embedded commands.
- Capability inventory: The skill executes scripts/generate_image.py which performs network requests to the Gemini API and writes image files to the local file system.
- Sanitization: No input filtering, escaping, or sanitization of the prompt is described.
- [EXTERNAL_DOWNLOADS]: The skill's metadata contains a URL pointing to an external source repository not associated with the vendor.
- Evidence: The url field in the SKILL.md frontmatter points to github.com/sickn33/antigravity-awesome-skills.
Audit Metadata