The Agent Skills Directory

Authentication via Azure CLI: The skill utilizes az account get-access-token to retrieve bearer tokens for interacting with the Dataverse API. This is a standard method for programmatic access within the Azure ecosystem and relies on the user's existing CLI session.\n- Management of Dataverse Schema: The skill includes functions to create and modify Dataverse tables and columns using REST API calls. These operations are intended for environment customization and require high-privilege security roles, such as System Administrator.\n- Processing of User-Supplied Schema Definitions: The skill accepts user input to define new Dataverse tables and relationships. This represents an indirect prompt injection surface where external data influences system operations.\n
Ingestion points: User-provided table names and descriptions in Step 1 of the planning workflow.\n
Boundary markers: The prompt instructions do not explicitly define delimiters for user-supplied content during interpolation.\n
Capability inventory: The skill utilizes Invoke-RestMethod (POST, PATCH) to modify Dataverse schema and the pa CLI to add data sources.\n
Sanitization: While the skill uses ConvertTo-Json for API payloads, it does not demonstrate explicit validation or sanitization of user-provided strings against injection attacks.\n- Generation and Compilation of Code: The skill generates TypeScript models and services based on Dataverse definitions and performs a build step using standard package management tools. This follows common development practices for code generation and environment setup.

add-dataverse