feat(gooddata-sdk): [AUTO] Add resolveLlmProviders endpoint and ResolvedLlm polymorphic schemas

[yenkins-admin]

Summary

Added SDK wrapper for the new resolveLlmProviders GET endpoint (/api/v1/actions/workspaces/{workspaceId}/ai/resolveLlmProviders). Created four new model classes (CatalogResolvedLlms, CatalogResolvedLlmEndpoint, CatalogResolvedLlmProvider, CatalogResolvedLlmModelItem) in a new entity_model file, added resolve_llm_providers() service method to CatalogWorkspaceContentService, exported all new classes from gooddata_sdk/init.py, and added pure-unit tests covering the polymorphic parsing logic.

Impact: new_feature | Services: gooddata-afm-client

Files changed

• packages/gooddata-sdk/src/gooddata_sdk/catalog/workspace/entity_model/resolved_llm.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_resolved_llm.py

Agent decisions

Decisions (4)

Service method placement — Added resolve_llm_providers() to CatalogWorkspaceContentService

• Alternatives: Add to CatalogWorkspaceService (workspace management), Add to CatalogOrganizationService alongside LLM provider CRUD
• Why: The endpoint is workspace-scoped (/workspaces/{id}/ai/resolveLlmProviders), matching other workspace content action methods already in this service (compute_valid_objects, get_label_elements, get_dependent_entities_graph).

Polymorphic oneOf parsing (endpoint vs provider) — Check for presence of the 'models' key in the raw data dict to dispatch to CatalogResolvedLlmProvider vs CatalogResolvedLlmEndpoint

• Alternatives: Use a discriminator field (none exists), Merge both subtypes into a single class with optional models field
• Why: No discriminator field exists in the spec. The only structural difference between the two subtypes is the presence of 'models' in ResolvedLlmProvider. Consistent with how llm_provider.py handles similar type detection.

client_class() with deferred imports — Lazy-import generated client classes inside client_class() static methods to avoid circular imports at module load time

• Alternatives: Top-level imports from gooddata_api_client
• Why: Matches existing pattern in graph.py and other entity_model files.

Test type: unit vs integration — Pure unit tests (no mocks, no cassettes)

• Alternatives: Integration test with VCR cassette (unnecessary for pure parsing logic)
• Why: The only non-trivial logic is dict-based type dispatch in CatalogResolvedLlms.from_api(). Unit tests cover all branching paths with zero external dependencies.

Assumptions to verify (3)

• The API returns 'models' only for provider-backed configurations and never for endpoint-backed ones, making the 'models' key a reliable discriminator.
• When _check_return_type=False, the actions_api.resolve_llm_providers() return value supports dict-style .get() access (consistent with how other _check_return_type=False calls are consumed in the codebase).
• The 'data' field is absent or null when no LLM is configured, rather than raising an HTTP error.

Risks (2)

• If the live backend returns a 404 or non-200 when no LLM is configured (instead of {data: null}), the service method will raise rather than returning CatalogResolvedLlms(data=None).
• If the API response dict uses a key other than 'models' the type-dispatch heuristic would break silently (parsing as endpoint instead of provider).

Layers touched (3)

• entity_model — New file with CatalogResolvedLlms, CatalogResolvedLlmEndpoint, CatalogResolvedLlmProvider, CatalogResolvedLlmModelItem
  • packages/gooddata-sdk/src/gooddata_sdk/catalog/workspace/entity_model/resolved_llm.py
• public_api — Added resolve_llm_providers() to CatalogWorkspaceContentService; exported 4 new classes from init.py
  • packages/gooddata-sdk/src/gooddata_sdk/catalog/workspace/content_service.py
  • packages/gooddata-sdk/src/gooddata_sdk/__init__.py
• tests — Unit tests for polymorphic from_api() parsing — no cassettes needed
  • packages/gooddata-sdk/tests/catalog/test_resolved_llm.py

Source commits (gdc-nas)

• 699c75c Merge pull request #21420 from hkad98/jkd/resolve-llm-provider

OpenAPI diff

diff --git a/microservices/afm-exec-api/src/test/resources/openapi/open-api-spec.json
@@ -3825,20 +3825,38 @@
-      "ResolvedLlmEndpoint" : {
+      "ResolvedLlm" : {
+        "description" : "The resolved LLM configuration, or null if none is configured.",
         "properties" : {
           "id" : { "type" : "string" },
           "title" : { "type" : "string" }
         },
         "required" : [ "id", "title" ],
         "type" : "object"
       },
+      "ResolvedLlmEndpoint" : {
+        "allOf" : [ { "$ref" : "#/components/schemas/ResolvedLlm" }, {
+          "properties" : {
+            "id" : { "description" : "Endpoint Id", "type" : "string" },
+            "title" : { "description" : "Endpoint Title", "type" : "string" }
+          }, "type" : "object"
+        } ],
+        "required" : [ "id", "title" ],
+        "type" : "object"
+      },
+      "ResolvedLlmProvider" : {
+        "allOf" : [ { "$ref" : "#/components/schemas/ResolvedLlm" }, {
+          "properties" : {
+            "id" : { "description" : "Provider Id", "type" : "string" },
+            "models" : { "items" : { "$ref" : "#/components/schemas/LlmModel" }, "maxItems" : 1, "minItems" : 1, "type" : "array" },
+            "title" : { "description" : "Provider Title", "type" : "string" }
+          }, "type" : "object"
+        } ],
+        "required" : [ "id", "models", "title" ],
+        "type" : "object"
+      },
+      "ResolvedLlms" : {
+        "properties" : {
+          "data" : { "oneOf" : [ { "$ref" : "#/components/schemas/ResolvedLlmEndpoint" }, { "$ref" : "#/components/schemas/ResolvedLlmProvider" } ] }
+        },
+        "type" : "object"
+      },
@@ -5987,6 +6044,36 @@
+    "/api/v1/actions/workspaces/{workspaceId}/ai/resolveLlmProviders" : {
+      "get" : {
+        "description" : "Resolves the active LLM configuration for the given workspace.",
+        "operationId" : "resolveLlmProviders",
+        "parameters" : [ { "description" : "Workspace identifier", "in" : "path", "name" : "workspaceId", "required" : true, "schema" : { "pattern" : "^(?!\\.)[.A-Za-z0-9_-]{1,255}$", "type" : "string" } } ],
+        "responses" : { "200" : { "content" : { "application/json" : { "schema" : { "$ref" : "#/components/schemas/ResolvedLlms" } } }, "description" : "OK" } },
+        "summary" : "Get Active LLM configuration for this workspace",
+        "tags" : [ "Smart Functions", "actions" ]
+      }
+    },

Workflow run

---

Generated by SDK OpenAPI Sync workflow

[codecov]

Codecov Report

❌ Patch coverage is 83.33333% with 10 lines in your changes missing coverage. Please review.
✅ Project coverage is 78.68%. Comparing base (37d0593) to head (7393758).

Files with missing lines	Patch %	Lines
...sdk/catalog/workspace/entity_model/resolved_llm.py	85.45%	8 Missing ⚠️
.../gooddata_sdk/catalog/workspace/content_service.py	50.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1554      +/-   ##
==========================================
+ Coverage   78.66%   78.68%   +0.01%     
==========================================
  Files         230      231       +1     
  Lines       15405    15465      +60     
==========================================
+ Hits        12118    12168      +50     
- Misses       3287     3297      +10     

☔ 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.