databasehub-cli
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [Data Ingestion Surface]: The skill is designed to retrieve and analyze data from the Microsoft Fabric API. Processing content from external systems is a standard security consideration for AI agents.\n
- Ingestion points: Data enters the agent context through responses from various Database Hub API routes documented in
references/consumption.mdandreferences/operations.md.\n - Boundary markers: The instructions are absent regarding the use of specific delimiters or 'ignore embedded instructions' warnings for the API-returned content.\n
- Capability inventory: As detailed in
references/fabric-api-basics.mdandreferences/operations.md, the skill executes shell commands viaaz restand performs local file operations (write and delete) using PowerShell to manage request payloads.\n - Sanitization: The skill contains instructions in
SKILL.mdto report sanitizederrors, but explicit validation or filtering logic for the primary data payload is not provided.
Audit Metadata