Support Snowflake Cortex as one of the Runtime

[sgsshankar]

Prerequisites

• I have checked the latest documentationand this feature doesn't exist
• I have searched for similar feature requests and found none
• I have linked to related issues or discussions in the description (if any exist)

Type of Improvement

New Functionality

Proposed Solution

Support running Azure OpenAI and other models in Snowflake Cortex as one of the Runtime for Prompty. It supports various Authentication mechanisms with Snowflake as well as the Cortex COMPLETE function similar to the standard Prompty Specification.

PR for the Feature is #246

[sethjuarez]

Closing this as the v2 architecture already supports adding new providers as plugins. Reasoning (from Copilot, reviewing on behalf of @sethjuarez):

Prompty v2 has a pluggable provider system. Adding Snowflake Cortex doesn't require any changes to the core runtime. You can build a standalone package (e.g. prompty-snowflake) that implements an executor and processor, then registers them via entry points (Python), the registry API (TypeScript/C#), or a plugin package.

How to add a new provider in v2:

1. Create a package with an executor (sends messages to the Cortex API) and a processor (extracts content from the response)
2. Register via entry points (Python) or registerExecutor/registerProcessor (TS) or InvokerRegistry.RegisterExecutor (C#)
3. Set model.provider: "snowflake" in .prompty files

See the provider documentation and the existing OpenAI/Azure/Anthropic providers as reference implementations.

The v1 PR (#246) is stale — the runtime has been completely rebuilt for v2. A new implementation against the v2 API would be straightforward.

We'd welcome a community contribution for a Snowflake Cortex provider package!