feat(azure_foundry): auto-discover deployed models on Azure AI Services#303
Open
breedx wants to merge 5 commits intompfaffenberger:mainfrom
Open
feat(azure_foundry): auto-discover deployed models on Azure AI Services#303breedx wants to merge 5 commits intompfaffenberger:mainfrom
breedx wants to merge 5 commits intompfaffenberger:mainfrom
Conversation
/foundry-setup now queries the Azure Management API to find all deployments on the specified AI Services account. Supports both Anthropic and OpenAI model formats automatically. - Add discovery.py: find_account() and list_deployments() via ARM API - Add azure_foundry_openai model type for OpenAI on Azure with AD auth - Setup wizard: auto-discover → confirm → configure, manual fallback - remove/get/status commands now handle both model types
Author
|
- Discovery: find_account, list_deployments, management token/API mocks - add_discovered_models_to_config: OpenAI, Anthropic, and mixed deployments - remove: verifies both azure_foundry and azure_foundry_openai types cleaned - OpenAI model handler: no resource, no name, auth failure, success - Updated register_model_types test for 2 types
Owner
|
Can we get on a super quick call and you can give me a demo? :D |
Author
I sent you a teams meetings request for tomorrow, if that is ok :) |
Discovery was hardcoding 128K for all OpenAI models. Added a lookup table since Azure's catalog API doesn't expose context window sizes.
Author
|
Author
had to set the thinking to 'detailed'
|
Author
|
codex works |
- Use OpenAIResponsesModel for GPT-5 models on Azure Foundry - Add azure_foundry_openai to uses_responses_api check - Default reasoning_summary to "detailed" (auto breaks 5.3 codex) - Auto-discover supported_settings per model family (GPT-5 gets reasoning_effort, summary, verbosity) - Update tests for Responses API model path
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
/foundry-setupnow auto-discovers all deployed models on an Azure AI Services account via the Azure Management API. Supports both Anthropic and OpenAI model formats automatically.Previously, setup required manually entering deployment names for three hardcoded Claude tiers. Now it queries the account, shows what's deployed, and configures everything in one step.
What changed
discovery.py(new) —find_account()searches across subscriptions,list_deployments()enumerates deployments via ARM APIregister_callbacks.py— setup wizard uses auto-discovery with manual fallback; newazure_foundry_openaimodel type handler for OpenAI models with Azure AD token authutils.py—add_discovered_models_to_config()classifies Anthropic vs OpenAI; remove/get handle both typesprovider_identity.py— +1 line: registerazure_foundry_openaiidentityHow it works
/foundry-setup, enters resource nameTest plan
/foundry-setupwith valid Azure AI Services resource — discovers all deployments/foundry-statusshows both model types/foundry-removecleans up both types