feat(gooddata-sdk): [AUTO] Remove LLM endpoint schemas and replace with 410 Gone

[yenkins-admin]

Summary

Added CatalogResolvedLlmProvider model and resolve_llm_providers() method to CatalogWorkspaceContentService as the SDK-level replacement for the now-removed resolveLlmEndpoints endpoint (which now returns 410 Gone). The old LlmEndpoint-based endpoints were never wrapped in the SDK wrapper layer; this change adds the new resolveLlmProviders action wrapper and exports the result type publicly.

Impact: modification | Services: gooddata-afm-client

Source commits (gdc-nas):

• 0d015ba by Jan Kadlec — Merge pull request #22495 from hkad98/jkd/llm-entity-removal

Files changed

• packages/gooddata-sdk/src/gooddata_sdk/catalog/organization/entity_model/llm_provider.py
• packages/gooddata-sdk/src/gooddata_sdk/catalog/workspace/content_service.py
• packages/gooddata-sdk/src/gooddata_sdk/__init__.py
• packages/gooddata-sdk/tests/catalog/test_catalog_workspace_content.py

Agent decisions

Decisions (3)

where to add resolve_llm_providers method — CatalogWorkspaceContentService in content_service.py

• Alternatives: ComputeService (handles AI workspace actions), CatalogOrganizationService (handles LLM provider CRUD)
• Why: resolve_llm_providers is a workspace-level configuration query. CatalogWorkspaceContentService already owns workspace-level action calls via _actions_api. ComputeService was the alternative but its AI methods are compute/inference operations, not configuration resolution.

no_changes vs new_feature for removed endpoints — Implement resolve_llm_providers as the replacement wrapper

• Alternatives: status=no_changes (the removed endpoints were never in the SDK wrapper)
• Why: The old resolveLlmEndpoints was never wrapped in the SDK. Per task instructions, when sdk_impact=modification but no integration test exists for the changed endpoint, treat as new_feature. The replacement resolveLlmProviders endpoint is the natural SDK addition.

CatalogResolvedLlmProvider model placement — Added to llm_provider.py in the organization entity_model directory

• Alternatives: Separate file in workspace/entity_model/, Inline dict return from resolve_llm_providers
• Why: The resolved provider is conceptually related to the LlmProvider domain. Co-locating it with other LLM provider models keeps the domain coherent. It reuses CatalogLlmProviderModel for the models list.

Assumptions to verify (3)

• The ty check errors in execution.py (pyarrow unresolved import) and filter.py (EmptyValueHandling return type) are pre-existing and not caused by this PR.
• resolve_llm_providers returns None when the workspace has no LLM provider configured (data field absent or null in response).
• CatalogLlmProviderModel.id and .family are always non-empty strings per the OpenAPI spec (both are required fields in LlmModel).

Risks (2)

• ResolvedLlms.data is a oneOf[ResolvedLlmEndpoint, ResolvedLlmProvider] in the current generated client. After the api-client is regenerated to match the new spec (removing ResolvedLlmEndpoint), the discriminator logic changes. The SDK's from_api uses safeget which is type-agnostic, so it should continue to work.
• The response field 'data' could be missing entirely when no provider is configured; the safeget None-guard handles this.

Layers touched (3)

• entity_model — Added CatalogResolvedLlmProvider read-only response model.
  • packages/gooddata-sdk/src/gooddata_sdk/catalog/organization/entity_model/llm_provider.py
• public_api — Added resolve_llm_providers() to CatalogWorkspaceContentService and exported CatalogResolvedLlmProvider from init.py.
  • packages/gooddata-sdk/src/gooddata_sdk/__init__.py
  • packages/gooddata-sdk/src/gooddata_sdk/catalog/workspace/content_service.py
• tests — Added test_resolve_llm_providers_integration integration test.
  • packages/gooddata-sdk/tests/catalog/test_catalog_workspace_content.py

OpenAPI diff

--- a/gooddata-afm-client.json
+++ b/gooddata-afm-client.json
@@ -5429,45 +5528,6 @@
-      "ResolvedLlmEndpoint": {
-        "allOf": [
-          { "$ref": "#/components/schemas/ResolvedLlm" },
-          {
-            "properties": {
-              "id": { "description": "Endpoint Id", "type": "string" },
-              "title": { "description": "Endpoint Title", "type": "string" }
-            },
-            "type": "object"
-          }
-        ],
-        "required": [ "id", "title" ],
-        "type": "object"
-      },
-      "ResolvedLlmEndpoints": {
-        "properties": {
-          "data": {
-            "items": { "$ref": "#/components/schemas/ResolvedLlmEndpoint" },
-            "type": "array"
-          }
-        },
-        "required": [ "data" ],
-        "type": "object"
-      },
-      "ValidateLLMEndpointByIdRequest": { ... },
-      "ValidateLLMEndpointRequest": { ... },
-      "ValidateLLMEndpointResponse": { ... },
@@ -6920,31 +6912,14 @@
     "/api/v1/actions/ai/llmEndpoint/test": {
       "post": {
         "deprecated": true,
-        "description": "Will be soon removed and replaced by testLlmProvider.",
+        "description": "Permanently removed. Use POST /api/v1/actions/ai/llmProvider/test instead. Always returns 410 Gone.",
         "responses": {
-          "200": { ... }
+          "410": { "description": "Gone" }
         },
-        "summary": "Validate LLM Endpoint",
+        "summary": "Validate LLM Endpoint (Removed)",
@@ -7654,7 +7613,7 @@
     "/api/v1/actions/workspaces/{workspaceId}/ai/resolveLlmEndpoints": {
       "get": {
         "deprecated": true,
-        "description": "Will be soon removed and replaced by LlmProvider-based resolution.",
+        "description": "Permanently removed. Use GET /api/v1/actions/workspaces/{workspaceId}/ai/resolveLlmProviders instead. Always returns 410 Gone.",
         "responses": {
-          "200": { ... }
+          "410": { "description": "Gone" }
         },
-        "summary": "Get Active LLM Endpoints for this workspace",
+        "summary": "Get Active LLM Endpoints for this workspace (Removed)",

Workflow run

---

Generated by SDK OpenAPI Sync workflow

[codecov]

Codecov Report

❌ Patch coverage is 83.33333% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 79.01%. Comparing base (38b0798) to head (cdf837c).

Files with missing lines	Patch %	Lines
.../catalog/organization/entity_model/llm_provider.py	80.00%	2 Missing ⚠️
.../gooddata_sdk/catalog/workspace/content_service.py	87.50%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1620   +/-   ##
=======================================
  Coverage   79.01%   79.01%           
=======================================
  Files         231      231           
  Lines       15619    15635   +16     
=======================================
+ Hits        12341    12354   +13     
- Misses       3278     3281    +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.