Let Codex route to Azure OpenAI and Microsoft Foundry Model Provider Services - #685
Merged
Merged
Conversation
masonc08
changed the title
Collaborator
|
can you show an e2e of codex talking to an azure open ai + foundry mps? ideally video |
masonc08
force-pushed
the
masonc08/codex-azure-openai
branch
2 times, most recently
from
September 17, 2026 15:49
46fb7fb to
caaf241
Compare
masonc08
added a commit
that referenced
this pull request
Sep 17, 2026
Cover routing Codex through an azure_openai Model Provider Service, backed by main.ucode.ci_azure_openai_mps (allowlist: gpt-5-nano, gpt-5.6-luna reasoning; gpt-4.1-mini non-reasoning). - e2e TestAzureOpenAiMps: direct Responses-API probes through the gateway MPS (no agent binary; gateway shard, independent of #685) — reasoning-capable models route and are switchable, a non-reasoning allow-listed model rejects reasoning.effort (parity: what fails natively fails here), the same model serves without reasoning, and an off-allowlist model is refused. - integration test_ug_configure_codex_azure_openai_mps: the real ug CLI picks the Azure MPS in the configure picker and completes a file task on gpt-5-nano. Requires #685 (adds azure_openai to Codex's provider-type allowlist). - run_integration.py / conftest.py: --codex-azure-provider(-model) plumbing. Co-authored-by: Isaac <no-reply@databricks.com>
masonc08
enabled auto-merge (squash)
September 17, 2026 18:57
Codex speaks the OpenAI dialect, which the gateway also fronts for
azure_openai-backed MPSes, but the provider-type allowlist only listed
`openai` — so launching Codex against an Azure OpenAI MPS was rejected
("codex can't route to (supported: openai)"). Add azure_openai to codex's
allowed provider types.
Co-authored-by: Isaac <no-reply@databricks.com>
microsoft_foundry speaks the same OpenAI dialect as azure_openai, so add it to codex's provider-type allowlist alongside it. Co-authored-by: Isaac <no-reply@databricks.com>
masonc08
force-pushed
the
masonc08/codex-azure-openai
branch
from
September 18, 2026 14:30
cd9ec9d to
b67a3c9
Compare
masonc08
added a commit
that referenced
this pull request
Sep 18, 2026
Cover routing Codex through an azure_openai Model Provider Service, backed by main.ucode.ci_azure_openai_mps (allowlist: gpt-5-nano, gpt-5.6-luna reasoning; gpt-4.1-mini non-reasoning). - e2e TestAzureOpenAiMps: direct Responses-API probes through the gateway MPS (no agent binary; gateway shard, independent of #685) — reasoning-capable models route and are switchable, a non-reasoning allow-listed model rejects reasoning.effort (parity: what fails natively fails here), the same model serves without reasoning, and an off-allowlist model is refused. - integration test_ug_configure_codex_azure_openai_mps: the real ug CLI picks the Azure MPS in the configure picker and completes a file task on gpt-5-nano. Requires #685 (adds azure_openai to Codex's provider-type allowlist). - run_integration.py / conftest.py: --codex-azure-provider(-model) plumbing. Co-authored-by: Isaac <no-reply@databricks.com>
masonc08
added a commit
that referenced
this pull request
Sep 18, 2026
Foundry speaks the same OpenAI dialect as Azure OpenAI, so extract the configure-picker + launch CUJ into a shared helper and add a thin test_ug_configure_codex_foundry_mps wrapper reusing it. Mirror the Azure flags/fixtures as --codex-foundry-provider / --codex-foundry-provider-model. Requires #685 (adds microsoft_foundry to Codex's allowlist) and a CI Foundry MPS. Co-authored-by: Isaac <no-reply@databricks.com>
rohita5l
approved these changes
Sep 18, 2026
masonc08
added a commit
that referenced
this pull request
Sep 18, 2026
Cover routing Codex through an azure_openai Model Provider Service, backed by main.ucode.ci_azure_openai_mps (allowlist: gpt-5-nano, gpt-5.6-luna reasoning; gpt-4.1-mini non-reasoning). - e2e TestAzureOpenAiMps: direct Responses-API probes through the gateway MPS (no agent binary; gateway shard, independent of #685) — reasoning-capable models route and are switchable, a non-reasoning allow-listed model rejects reasoning.effort (parity: what fails natively fails here), the same model serves without reasoning, and an off-allowlist model is refused. - integration test_ug_configure_codex_azure_openai_mps: the real ug CLI picks the Azure MPS in the configure picker and completes a file task on gpt-5-nano. Requires #685 (adds azure_openai to Codex's provider-type allowlist). - run_integration.py / conftest.py: --codex-azure-provider(-model) plumbing. Co-authored-by: Isaac <no-reply@databricks.com>
masonc08
added a commit
that referenced
this pull request
Sep 18, 2026
Foundry speaks the same OpenAI dialect as Azure OpenAI, so extract the configure-picker + launch CUJ into a shared helper and add a thin test_ug_configure_codex_foundry_mps wrapper reusing it. Mirror the Azure flags/fixtures as --codex-foundry-provider / --codex-foundry-provider-model. Requires #685 (adds microsoft_foundry to Codex's allowlist) and a CI Foundry MPS. Co-authored-by: Isaac <no-reply@databricks.com>
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
3 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.
What did you change, and why?
Change: Add
azure_openaiandmicrosoft_foundryto Codex's allowed provider types (_TOOL_PROVIDER_TYPES).Why: Codex speaks the OpenAI dialect, and the AI Gateway fronts
azure_openai- andmicrosoft_foundry-backed Model Provider Services with that same OpenAI surface — but the allowlist only listedopenai, so pointing Codex at one (via--provideror a managed config) failed with "codex can't route to (supported: openai)". This permits both.How do you know it works?
Screen.Recording.2026-09-17.at.14.27.01.mov
Screen.Recording.2026-09-17.at.14.55.59.mov
Testing: Extended the provider-support unit test to assert Codex accepts
azure_openaiandmicrosoft_foundry;test_managed_setup.py+test_databricks.pypass (369 tests). End-to-end routing against live Azure OpenAI / Foundry MPSes is being validated separately.This pull request and its description were written by Isaac.