sqldb-consumption-cli

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFE
Full Analysis
  • Prompt-Level Security Controls: The skill includes explicit instructions to the agent to refuse any commands that modify data (DML) or schema (DDL). This 'read-only' constraint is a proactive safety measure to prevent unauthorized changes to the database environment.
  • Secure Dynamic Query Generation: In the discovery reference file, the skill employs dynamic T-SQL to search across multiple tables. This is handled securely by using QUOTENAME() to escape database identifiers and using parameters for search values, which effectively prevents SQL injection vulnerabilities.
  • Data Ingestion Analysis: The skill processes database content (Ingestion Point: SQL query results). While this creates a potential surface for indirect prompt injection, the risk is mitigated by the skill's strict read-only mandate (Boundary Marker: read-only instruction) and its focus on metadata. The skill's capabilities are limited to standard database querying tools (Capability Inventory: sqlcmd, az CLI).
  • Trusted Tool Integration: The skill utilizes official Microsoft tools and standard utilities. The connection and authentication patterns follow established best practices, such as leveraging DefaultAzureCredential for Entra ID authentication.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 03:27 AM
Security Audit — agent-trust-hub — sqldb-consumption-cli