mdr-745-specialist
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves analyzing external medical device documentation, clinical evidence, and technical files provided by the user. This creates a vulnerability surface where malicious instructions hidden in these documents could influence the agent's behavior or outcomes. * Ingestion points: Technical documentation (Annex II & III), clinical evaluation data, and device descriptions. * Boundary markers: There are no instructions to use delimiters or ignore embedded prompts when processing user-provided documentation. * Capability inventory: The skill utilizes scripts (e.g., mdr-gap-analysis.py) for automated assessment and reporting based on external inputs. * Sanitization: No input validation or filtering mechanisms are described for the content being processed.
Audit Metadata