device-guidance-v1-0-2
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection Surface: The skill retrieves device information from an external tool and presents it to the user. Instructions within the skill specify how to map user selections (e.g., Option A) back to internal IDs for follow-up actions.\n
- Ingestion points: Data enters the context via the
Get Employee Assetstool output, which originates from a SharePoint source.\n - Boundary markers: The skill does not explicitly define delimiters or instructions to ignore embedded commands within the fetched asset data, though it uses a lettered mapping system (A, B, C) to separate user input from internal IDs.\n
- Capability inventory: The skill uses the
Get Employee Assetstool and implies a follow-up request action if an option is selected.\n - Sanitization: There are no explicit instructions for sanitizing or escaping the content retrieved from the external source.\n- Data Query Configuration: The instructions include specific mappings for internal field names (e.g.,
field_1for Status) to be used in OData filters. This is an operational requirement for the tool to function correctly and interact with the data source schema.
Audit Metadata