feat(gooddata-sdk): [AUTO] Add conversation responses list and feedback endpoints in gen-ai

[yenkins-admin]

Summary

Added SDK support for the two new gen-ai conversation response endpoints: GET conversation responses and PATCH response feedback. Created model classes (ConversationFeedback, ConversationTurnResponse, ConversationResponseList) in compute/model/ai_chat.py, added _do_get_request and _do_patch_request to GoodDataApiClient, added get_conversation_responses and set_conversation_response_feedback service methods to ComputeService, exported new classes from init.py, and added 12 unit tests covering all model serialization paths.

Impact: new_feature | Services: gen-ai
Tickets: GDAI-1448

Files changed

• packages/gooddata-sdk/src/gooddata_sdk/compute/model/ai_chat.py
• packages/gooddata-sdk/src/gooddata_sdk/client.py
• packages/gooddata-sdk/src/gooddata_sdk/compute/service.py
• packages/gooddata-sdk/src/gooddata_sdk/__init__.py
• packages/gooddata-sdk/tests/compute/test_ai_chat_models.py

Agent decisions

Decisions (4)

Model class location — compute/model/ai_chat.py — standalone attrs classes without Base

• Alternatives: catalog/workspace/entity_model/ with Base class, Inline dicts returned directly from service methods
• Why: These DTOs come from the gen-ai microservice action endpoints, not from the JSONAPI entity layer. Existing AI chat methods in ComputeService also don't wrap responses in Catalog* classes. Using compute/model/ matches the existing pattern for compute-adjacent models.

HTTP client methods — Added _do_get_request and _do_patch_request to GoodDataApiClient; also updated do_request to handle PATCH

• Alternatives: Use requests directly in the service layer without client methods, Extend actions_api by monkey-patching
• Why: The new endpoints (GET + PATCH) are not in the auto-generated API client. Mirroring the existing _do_post_request pattern keeps the client layer consistent. HttpMethod.PATCH was already defined but unsupported — wiring it in do_request is a natural extension.

ResponseFeedbackRequest body structure — Wrapped FeedbackDto in { 'feedback': {...} } envelope

• Alternatives: Send FeedbackDto fields (type, text) directly at the top level
• Why: The diff elides the full ResponseFeedbackRequest schema. A wrapper envelope is the most common pattern for PATCH request bodies, and it mirrors how ConversationTurnResponseDto contains a nested feedback field.

ConversationTurnResponse unused import fix — Ruff removed unused ConversationTurnResponse import from service.py automatically

• Alternatives: Keep import for user convenience
• Why: ConversationTurnResponse is only referenced via ConversationResponseList.from_api internally; no need to re-import it in the service layer. Ruff correctly identified and removed the redundant import.

Assumptions to verify (4)

• ResponseFeedbackRequest body wraps FeedbackDto as { 'feedback': { 'type': ..., 'text': ... } } — actual schema was elided in the diff as '{ ... }'
• ResponseFeedbackDto (also elided) is not needed for the SDK layer; only the request direction is relevant for the PATCH method
• The gen-ai endpoints use the same hostname as the rest of the GoodData API, differing only in path (/api/v1/ai/... vs /api/v1/actions/...)
• Bearer token auth is sufficient for these new endpoints (same as other do*_request methods)

Risks (2)

• Integration test for set_conversation_response_feedback may fail if ResponseFeedbackRequest actually has a flat structure (type/text at top level) rather than a nested feedback object
• get_conversation_responses may fail if the gen-ai service is not deployed at the same host/port as the main API in the test environment

Layers touched (4)

• entity_model — New model classes for ConversationFeedback, ConversationTurnResponse, ConversationResponseList — not JSONAPI entities, so no Base class inheritance
  • packages/gooddata-sdk/src/gooddata_sdk/compute/model/ai_chat.py
• public_api — Exported ConversationFeedback, ConversationTurnResponse, ConversationResponseList
  • packages/gooddata-sdk/src/gooddata_sdk/__init__.py
• converter — Added _do_get_request/_do_patch_request to client; added get_conversation_responses and set_conversation_response_feedback to ComputeService
  • packages/gooddata-sdk/src/gooddata_sdk/client.py
  • packages/gooddata-sdk/src/gooddata_sdk/compute/service.py
• tests — 12 unit tests covering from_api, to_api_dict, and roundtrip for all three model classes
  • packages/gooddata-sdk/tests/compute/test_ai_chat_models.py

Source commits (gdc-nas)

• e58ab04 Merge pull request #21392 from gooddata/js/GDAI-1448

OpenAPI diff

diff --git a/microservices/gen-ai/src/test/resources/openapi/open-api-spec.json b/microservices/gen-ai/src/test/resources/openapi/open-api-spec.json
@@ -261,6 +261,62 @@
+      "ConversationResponseListDto": {
+        "additionalProperties": false,
+        "properties": {
+          "responses": {
+            "description": "Conversation responses.",
+            "items": { "$ref": "#/components/schemas/ConversationTurnResponseDto" },
+            "title": "Responses",
+            "type": "array"
+          }
+        },
+        "required": ["responses"],
+        "title": "ConversationResponseListDto",
+        "type": "object"
+      },
+      "ConversationTurnResponseDto": {
+        "properties": {
+          "createdAt": { "description": "Response creation timestamp (ISO-8601 UTC).", "type": "string" },
+          "feedback": { "anyOf": [ { "$ref": "#/components/schemas/FeedbackDto" }, { "type": "null" } ] },
+          "responseId": { "description": "Response identifier.", "type": "string" },
+          "updatedAt": { "description": "Response update timestamp (ISO-8601 UTC).", "type": "string" }
+        },
+        "required": ["responseId", "createdAt", "updatedAt"],
+        "type": "object"
+      },
+      "FeedbackDto": {
+        "properties": {
+          "text": { "anyOf": [ { "type": "string" }, { "type": "null" } ], "description": "Optional free-form feedback comment." },
+          "type": { "description": "Feedback type.", "enum": ["POSITIVE", "NEGATIVE"], "type": "string" }
+        },
+        "required": ["type"],
+        "type": "object"
+      },
+      "ResponseFeedbackDto": { ... },
+      "ResponseFeedbackRequest": { ... },
@@ -1598,6 +1722,126 @@
+    "/api/v1/ai/workspaces/{workspace_id}/chat/conversations/{conversation_id}/responses": {
+      "get": {
+        "operationId": "get_conversation_responses_api_v1_ai_workspaces__workspace_id__chat_conversations__conversation_id__responses_get",
+        "responses": { "200": { "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ConversationResponseListDto" } } }, "description": "Successful Response" } },
+        "summary": "Get Conversation Responses",
+        "tags": ["Responses"]
+      }
+    },
+    "/api/v1/ai/workspaces/{workspace_id}/chat/conversations/{conversation_id}/responses/{response_id}": {
+      "patch": {
+        "operationId": "patch_response_api_v1_ai_workspaces__workspace_id__chat_conversations__conversation_id__responses__response_id__patch",
+        "requestBody": { "content": { "application/json": { "schema": { "$ref": "#/components/schemas/ResponseFeedbackRequest" } } }, "required": true },
+        "responses": { "204": { "description": "Successful Response" } },
+        "summary": "Patch Response",
+        "tags": ["Responses"]
+      }
+    }

Workflow run

---

Generated by SDK OpenAPI Sync workflow

[codecov]

Codecov Report

❌ Patch coverage is 66.10169% with 20 lines in your changes missing coverage. Please review.
✅ Project coverage is 78.61%. Comparing base (37d0593) to head (4a33a49).

Files with missing lines	Patch %	Lines
packages/gooddata-sdk/src/gooddata_sdk/client.py	15.38%	11 Missing ⚠️
...s/gooddata-sdk/src/gooddata_sdk/compute/service.py	25.00%	9 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1551      +/-   ##
==========================================
- Coverage   78.66%   78.61%   -0.05%     
==========================================
  Files         230      231       +1     
  Lines       15405    15463      +58     
==========================================
+ Hits        12118    12157      +39     
- Misses       3287     3306      +19     

☔ 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.