opper-multimodal
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of external, potentially untrusted media files such as PDFs, images, and audio, which can contain hidden instructions designed to manipulate agent behavior. * Ingestion points: Data processed via /v3/ocr, /v3/audio/transcriptions, and vision-capable chat endpoints in SKILL.md. * Boundary markers: The instructions do not define specific delimiters or warnings to ignore embedded content. * Capability inventory: The skill utilizes network operations to interact with API endpoints and performs file management tasks. * Sanitization: No explicit sanitization or filtering logic is provided within the instruction set.
- [EXTERNAL_DOWNLOADS]: The skill implements a version-check protocol that retrieves metadata and updated instruction files from a remote source. * It fetches https://skills.opper.ai/versions.json to verify the local installation. * It instructs the agent to read and follow a remote SKILL.md from https://skills.opper.ai/opper-multimodal/SKILL.md if an update is detected. * These operations target the official vendor domain and represent standard maintenance behavior.
Audit Metadata