gemini
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for utilizing the
x geminiand@geminicommand-line interfaces to interact with Google Gemini AI models for tasks such as chat, file translation, and image analysis. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from local files (e.g., images, markdown files) using an LLM, which introduces a potential surface for indirect prompt injection. This is a common characteristic of AI integration tools and is considered a low risk in this context.
- Ingestion points: Local files passed via the
--fileargument inSKILL.md(e.g.,./pic.jpg,./abstract.en.md). - Boundary markers: Absent; the instructions do not specify the use of delimiters or "ignore" instructions for external content.
- Capability inventory: Uses the
x geminitool to communicate with external AI services. - Sanitization: Absent; there is no mention of content filtering or validation before processing.
Audit Metadata