blip-2-vision-language
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external image files and text prompts which serve as entry points for untrusted data into the agent's context.
- Ingestion points: Functions like "Image.open()" and "processor(images=image, text=question)" in "SKILL.md" and "advanced-usage.md" ingest data from external sources.
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the provided code snippets.
- Capability inventory: The model generates natural language output via "model.generate()", which is then printed to the console or integrated into downstream workflows like LangChain agents.
- Sanitization: No input sanitization or output filtering is implemented in the tutorial code.
Audit Metadata