The Agent Skills Directory

[Database Interaction Patterns]: The skill uses pyodbc and sqlcmd to manage the Azure SQL Database engine. It correctly employs parameterized queries (using ? placeholders) for inserting and querying data, which is a standard security practice to ensure that user-provided text is handled safely.
[Container Deployment]: The skill instructs the use of docker run to pull and execute an image from an Azure Container Registry (sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io). This is a routine procedure for setting up local development environments using specific vendor software images.
[Credential Management]: The scripts use a hardcoded placeholder password (YourStr0ng_Passw0rd) for the initial container setup. This is a common pattern in development documentation and is intended for a local, isolated instance.
[Local Network Services]: The embedding functionality is designed to interact with a local Ollama service (http://localhost:11434). This approach ensures that sensitive text data remains within the local network environment during the embedding process.
[Dynamic SQL Construction]: While the skill uses string interpolation to define the vector dimension (e.g., VECTOR({EMBED_DIM})), the dimension itself is defined as a static constant in the provided examples. This pattern is used here to address a specific T-SQL syntax requirement for the native VECTOR type.

azuresql-db-rag