LCORE-1880: Split requests into multiple files#1684
Conversation
WalkthroughThis PR reorganizes REST API request models from a monolithic ChangesModel Reorganization & Type Extraction
Sequence DiagramSequence diagrams are not applicable to this change, as it is primarily a refactoring of code organization and module structure without introducing new functional interactions or control flow patterns between components. Estimated code review effort🎯 4 (Complex) | ⏱️ ~50 minutes 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
✨ Simplify code
Warning Review ran into problems🔥 ProblemsGit: Failed to clone repository. Please run the Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 10
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
src/models/common/health.py (1)
19-39:⚠️ Potential issue | 🟡 Minor | ⚡ Quick win
ProviderHealthStatus.statusisstrdespite the co-locatedHealthStatusenum; docstring/examples reference a non-existent"unhealthy"value.Two related problems:
- Type:
status: strgives up the validation and IDE-completion benefits ofHealthStatus. Typing it asHealthStatus(once the casing fix above is applied) enforces the closed set of values.- Docs/examples mismatch: The class docstring (line 24) and field
examples(line 33) both list'unhealthy', which is not a member ofHealthStatus. The closest valid values areHEALTHY = "healthy"andERROR = "error".🛡️ Proposed fix
- status: str = Field( + status: HealthStatus = Field( description="The health status", - examples=["ok", "unhealthy", "not_implemented"], + examples=["ok", "error", "not_implemented"], )And update the docstring attribute line:
- status: The health status ('ok', 'unhealthy', 'not_implemented'). + status: The health status ('ok', 'error', 'not_implemented', 'healthy', 'unknown').🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/common/health.py` around lines 19 - 39, Change the ProviderHealthStatus.status field to use the HealthStatus enum instead of str (i.e., status: HealthStatus) so Pydantic validates the closed set; update the class docstring and the Field examples to remove the invalid 'unhealthy' value and use the actual enum values (e.g., "healthy" and "error" or whatever HealthStatus defines), and add an import for HealthStatus if not already present.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/openapi.json`:
- Line 11465: Regenerate docs/openapi.json from the canonical schema/models
using the project's documented generator command prefixed with "uv run" (e.g.,
run the repo's openapi/doc generation task via "uv run
<generate-openapi-command>"), replace the committed drifted docs/openapi.json
with the newly generated file, commit the updated artifact, and push; repeat the
same regeneration/commit for the other out-of-date generated entries referenced
in the review so the generated docs remain synchronized with the source models.
In `@src/models/api/requests/catalog.py`:
- Around line 8-20: The class docstring for ModelFilter is incorrect: update the
docstring so it no longer says "Required model type" and instead describes
model_type as an optional filter (e.g., "Optional model type filter to select
models such as 'llm', 'embeddings'"); ensure the Attributes section matches the
field definition model_type: Optional[str] and the existing Field description,
and keep model_config = {"extra": "forbid"} and the symbol names ModelFilter and
model_type unchanged.
In `@src/models/api/requests/feedback.py`:
- Around line 170-173: The presence validation currently treats whitespace-only
strings as non-empty because it checks self.user_feedback == ""; update the
check in the conditional that references self.sentiment, self.user_feedback and
self.categories to treat whitespace-only feedback as empty by trimming before
emptiness check (e.g., use self.user_feedback.strip() or a helper like
cleaned_feedback = (self.user_feedback or "").strip() and test that for
equality/emptiness) so that `" "` does not satisfy the “at least one feedback
form” rule; adjust the conditional that uses self.user_feedback accordingly.
In `@src/models/api/requests/mcp_servers.py`:
- Line 5: Replace usages of Optional[int] combined with Field(gt=0) in this
Pydantic model with Optional[PositiveInt] and update the imports accordingly:
add PositiveInt to the pydantic import line (replace or augment "from pydantic
import BaseModel, Field, field_validator" to include PositiveInt) and remove the
redundant gt=0 constraint from those Field(...) calls; specifically update each
field currently typed as Optional[int] + Field(gt=0) (the occurrences around the
later model fields) to Optional[PositiveInt] so validation uses the annotated
type instead of manual gt checks.
In `@src/models/api/requests/responses_openai.py`:
- Around line 157-169: The mutual-exclusivity check and the SUID validator treat
empty string as missing; update them to use explicit None checks so empty string
is considered provided and validated. Change the mutual-exclusivity condition to
check "self.conversation is not None" and "self.previous_response_id is not
None" (the block that raises ValueError when both are present) and change the
`@field_validator` method check_suid to "if value is not None and not
suid.check_suid(value)" so empty string fails SUID validation rather than being
skipped.
In `@src/models/api/requests/vector_stores.py`:
- Line 3: Replace the "from typing import Any, Optional, Self" import so that
Self is imported from typing_extensions (use "from typing import Any, Optional"
plus "from typing_extensions import Self") in
src/models/api/requests/vector_stores.py and the sibling request files query.py,
feedback.py, and responses_openai.py; update the import lines that reference
Self (and any annotations using Self) to use typing_extensions.Self to satisfy
Ruff UP035 and the project coding guidelines.
In `@src/models/common/health.py`:
- Around line 9-16: The HealthStatus enum has an inconsistent value: change
HealthStatus.ERROR from "Error" to lowercase "error" so it matches the other
enum members; update the enum definition in class HealthStatus (the ERROR
member) and any tests or comparisons that rely on HealthStatus.ERROR.value to
ensure they expect the lowercase string.
In `@src/models/common/query.py`:
- Around line 13-59: Change Attachment's model_config to use Pydantic v2's typed
ConfigDict like SolrVectorSearchRequest for consistency: import ConfigDict from
pydantic and replace the raw dict assigned to Attachment.model_config with a
ConfigDict(...) containing the same keys ("extra" and "json_schema_extra" with
the examples). Keep the contents identical but use ConfigDict to match the style
used by SolrVectorSearchRequest.
In `@src/models/README.md`:
- Around line 15-25: Update the README headings to satisfy MD022 by inserting a
single blank line between each heading and its following description for all
sections; specifically edit the headings "## [api/](api/)", "##
[common/](common/)", "## [database/](database/)", "## [rlsapi/](rlsapi/)" and
the pre-existing top sections that follow the same pattern so each heading
(e.g., the lines containing those exact strings) is immediately followed by one
blank line before the descriptive text.
In `@tests/unit/models/requests/test_query_request.py`:
- Around line 95-109: Test currently only checks that QueryRequest accepts
"text/plain"; add cases to cover the validator's other branches: assert that
"application/json" (MEDIA_TYPE_JSON) is accepted by constructing QueryRequest
(class QueryRequest) with media_type="application/json" and asserting
qr.media_type, and assert that an invalid media_type (e.g., "image/png") raises
ValueError from the model-validator validate_media_type in
src/models/api/requests/query.py by using pytest.raises(ValueError) when
instantiating QueryRequest with that value; reference QueryRequest and the
constants MEDIA_TYPE_JSON and MEDIA_TYPE_TEXT in the test names/comments so it's
clear which branch is covered.
---
Outside diff comments:
In `@src/models/common/health.py`:
- Around line 19-39: Change the ProviderHealthStatus.status field to use the
HealthStatus enum instead of str (i.e., status: HealthStatus) so Pydantic
validates the closed set; update the class docstring and the Field examples to
remove the invalid 'unhealthy' value and use the actual enum values (e.g.,
"healthy" and "error" or whatever HealthStatus defines), and add an import for
HealthStatus if not already present.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro
Run ID: da100b50-94de-486d-be81-ed767c3d939e
📒 Files selected for processing (71)
AGENTS.mddocs/openapi.jsonsrc/app/endpoints/a2a.pysrc/app/endpoints/conversations_v1.pysrc/app/endpoints/conversations_v2.pysrc/app/endpoints/feedback.pysrc/app/endpoints/health.pysrc/app/endpoints/mcp_servers.pysrc/app/endpoints/models.pysrc/app/endpoints/prompts.pysrc/app/endpoints/query.pysrc/app/endpoints/responses.pysrc/app/endpoints/stream_interrupt.pysrc/app/endpoints/streaming_query.pysrc/app/endpoints/vector_stores.pysrc/models/README.mdsrc/models/api/requests/__init__.pysrc/models/api/requests/catalog.pysrc/models/api/requests/conversations.pysrc/models/api/requests/feedback.pysrc/models/api/requests/mcp_servers.pysrc/models/api/requests/prompts.pysrc/models/api/requests/query.pysrc/models/api/requests/responses_openai.pysrc/models/api/requests/vector_stores.pysrc/models/common/__init__.pysrc/models/common/feedback_category.pysrc/models/common/health.pysrc/models/common/query.pysrc/models/context.pysrc/models/requests.pysrc/utils/query.pysrc/utils/responses.pysrc/utils/shields.pysrc/utils/transcripts.pysrc/utils/vector_search.pytests/integration/endpoints/test_conversations_v1_integration.pytests/integration/endpoints/test_conversations_v2_integration.pytests/integration/endpoints/test_health_integration.pytests/integration/endpoints/test_model_list.pytests/integration/endpoints/test_query_byok_integration.pytests/integration/endpoints/test_query_integration.pytests/integration/endpoints/test_stream_interrupt_integration.pytests/integration/endpoints/test_streaming_query_byok_integration.pytests/integration/endpoints/test_streaming_query_integration.pytests/unit/app/endpoints/test_conversations.pytests/unit/app/endpoints/test_conversations_v2.pytests/unit/app/endpoints/test_feedback.pytests/unit/app/endpoints/test_health.pytests/unit/app/endpoints/test_mcp_servers.pytests/unit/app/endpoints/test_models.pytests/unit/app/endpoints/test_prompts.pytests/unit/app/endpoints/test_query.pytests/unit/app/endpoints/test_responses.pytests/unit/app/endpoints/test_responses_splunk.pytests/unit/app/endpoints/test_stream_interrupt.pytests/unit/app/endpoints/test_streaming_query.pytests/unit/app/endpoints/test_vector_stores.pytests/unit/models/requests/README.mdtests/unit/models/requests/__init__.pytests/unit/models/requests/test_attachment.pytests/unit/models/requests/test_feedback_request.pytests/unit/models/requests/test_feedback_status_update_request.pytests/unit/models/requests/test_query_request.pytests/unit/models/requests/test_responses_request.pytests/unit/models/requests/test_vector_store_requests.pytests/unit/utils/test_prompts.pytests/unit/utils/test_query.pytests/unit/utils/test_responses.pytests/unit/utils/test_transcripts.pytests/unit/utils/test_vector_search.py
💤 Files with no reviewable changes (1)
- src/models/requests.py
📜 Review details
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (9)
- GitHub Check: build-pr
- GitHub Check: Konflux kflux-prd-rh02 / lightspeed-stack-on-pull-request
- GitHub Check: E2E: server mode / ci / group 3
- GitHub Check: E2E: server mode / ci / group 1
- GitHub Check: E2E: library mode / ci / group 1
- GitHub Check: E2E: library mode / ci / group 2
- GitHub Check: E2E: library mode / ci / group 3
- GitHub Check: E2E: server mode / ci / group 2
- GitHub Check: E2E Tests for Lightspeed Evaluation job
🧰 Additional context used
📓 Path-based instructions (5)
tests/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Use pytest for all unit and integration tests, not unittest
Use
pytest-mockfor AsyncMock objects in testsUse marker
pytest.mark.asynciofor async testsUnit tests require 60% coverage, integration tests require 10% coverage
Files:
tests/integration/endpoints/test_stream_interrupt_integration.pytests/unit/models/requests/test_feedback_status_update_request.pytests/unit/models/requests/test_feedback_request.pytests/unit/models/requests/test_attachment.pytests/unit/models/requests/__init__.pytests/unit/app/endpoints/test_responses.pytests/unit/utils/test_responses.pytests/integration/endpoints/test_query_byok_integration.pytests/unit/app/endpoints/test_conversations_v2.pytests/unit/app/endpoints/test_query.pytests/unit/app/endpoints/test_conversations.pytests/integration/endpoints/test_streaming_query_byok_integration.pytests/unit/models/requests/test_query_request.pytests/integration/endpoints/test_streaming_query_integration.pytests/unit/app/endpoints/test_feedback.pytests/integration/endpoints/test_model_list.pytests/unit/utils/test_vector_search.pytests/unit/app/endpoints/test_stream_interrupt.pytests/unit/app/endpoints/test_prompts.pytests/unit/app/endpoints/test_mcp_servers.pytests/integration/endpoints/test_conversations_v1_integration.pytests/unit/utils/test_transcripts.pytests/unit/models/requests/test_vector_store_requests.pytests/unit/utils/test_query.pytests/integration/endpoints/test_query_integration.pytests/unit/models/requests/test_responses_request.pytests/integration/endpoints/test_conversations_v2_integration.pytests/unit/app/endpoints/test_models.pytests/unit/app/endpoints/test_streaming_query.pytests/unit/utils/test_prompts.pytests/unit/app/endpoints/test_responses_splunk.pytests/unit/app/endpoints/test_vector_stores.pytests/unit/app/endpoints/test_health.pytests/integration/endpoints/test_health_integration.py
src/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Use absolute imports for internal modules (e.g.,
from authentication import get_auth_dependency)Use
from fastapi import APIRouter, HTTPException, Request, status, Dependsfor FastAPI dependenciesUse
from llama_stack_client import AsyncLlamaStackClientfor Llama Stack importsCheck
constants.pyfor shared constants before defining new onesAll modules must start with descriptive docstrings explaining purpose
Use
logger = get_logger(__name__)fromlog.pyfor module loggingUse
Final[type]as type hint for all constantsUse
@field_validatorand@model_validatorfor custom validation in Pydantic modelsUse type hints
Optional[FilePath],PositiveInt,SecretStrin Pydantic modelsAll functions require docstrings with brief descriptions following Google Python docstring conventions
Use complete type annotations for function parameters and return types, including
typing_extensions.Self, modernstr | intsyntax, andOptional[Type]Use snake_case with descriptive, action-oriented function names (get_, validate_, check_)
Avoid in-place parameter modification anti-patterns; return new data structures instead
Use
async deffor I/O operations and external API callsHandle
APIConnectionErrorfrom Llama Stack in error handlingUse standard log levels:
logger.debug()for detailed diagnostic information,logger.info()for general information,logger.warning()for unexpected events,logger.error()for serious problemsAll classes require descriptive docstrings explaining purpose
Use PascalCase with descriptive names and standard suffixes:
Configurationfor config classes,Error/Exceptionfor custom exceptions,Resolverfor strategy pattern,Interfacefor abstract base classesPydantic configuration models must extend
ConfigurationBase, data models must extendBaseModelAbstract classes must use ABC with
@abstractmethoddecoratorsUse complete type annotations for all class attributes; avoid using
AnyFo...
Files:
src/models/api/requests/conversations.pysrc/models/api/requests/__init__.pysrc/app/endpoints/mcp_servers.pysrc/models/common/__init__.pysrc/app/endpoints/vector_stores.pysrc/models/common/feedback_category.pysrc/app/endpoints/a2a.pysrc/models/context.pysrc/models/api/requests/catalog.pysrc/app/endpoints/feedback.pysrc/models/common/query.pysrc/models/common/health.pysrc/models/api/requests/prompts.pysrc/app/endpoints/stream_interrupt.pysrc/app/endpoints/prompts.pysrc/utils/vector_search.pysrc/models/api/requests/feedback.pysrc/app/endpoints/query.pysrc/models/api/requests/responses_openai.pysrc/models/api/requests/mcp_servers.pysrc/models/api/requests/vector_stores.pysrc/app/endpoints/conversations_v2.pysrc/models/api/requests/query.pysrc/app/endpoints/responses.pysrc/utils/responses.pysrc/app/endpoints/conversations_v1.pysrc/utils/query.pysrc/utils/shields.pysrc/app/endpoints/health.pysrc/app/endpoints/models.pysrc/app/endpoints/streaming_query.pysrc/utils/transcripts.py
AGENTS.md
📄 CodeRabbit inference engine (CLAUDE.md)
Document agent implementations and their configurations in AGENTS.md
Files:
AGENTS.md
src/**/__init__.py
📄 CodeRabbit inference engine (AGENTS.md)
Package
__init__.pyfiles must contain brief package descriptions
Files:
src/models/api/requests/__init__.pysrc/models/common/__init__.py
src/app/endpoints/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Use FastAPI
HTTPExceptionwith appropriate status codes for API endpoint error handling
Files:
src/app/endpoints/mcp_servers.pysrc/app/endpoints/vector_stores.pysrc/app/endpoints/a2a.pysrc/app/endpoints/feedback.pysrc/app/endpoints/stream_interrupt.pysrc/app/endpoints/prompts.pysrc/app/endpoints/query.pysrc/app/endpoints/conversations_v2.pysrc/app/endpoints/responses.pysrc/app/endpoints/conversations_v1.pysrc/app/endpoints/health.pysrc/app/endpoints/models.pysrc/app/endpoints/streaming_query.py
🧠 Learnings (7)
📓 Common learnings
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Always check `pyproject.toml` for existing dependencies before adding new ones
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Always verify current library versions in `pyproject.toml` rather than assuming versions
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Define E2E test steps in `tests/e2e/features/steps/` and maintain test list in `tests/e2e/test_list.txt`
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Run `uv run make format` to auto-format code with black + ruff before completion
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Run `uv run make verify` to run all linters (black, pylint, pyright, ruff, docstyle, check-types) before completion
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Create unit tests for new code before completion
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Use `uv run` prefix for all commands and `uv sync --group dev --group llslibdev` for dependencies
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: PR titles MUST start with a JIRA issue key prefix from the allowed list: `LCORE-`, `RSPEED-`, `MGTM-`, `OLS-`, `RHDHPAI-`, `LEADS-`
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Follow existing code patterns in the module you're modifying
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Never commit secrets/keys; use environment variables for sensitive data
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T11:16:13.327Z
Learning: Check `pyproject.toml` for Python versions and required dependencies before development
📚 Learning: 2026-01-12T10:58:40.230Z
Learnt from: blublinsky
Repo: lightspeed-core/lightspeed-stack PR: 972
File: src/models/config.py:459-513
Timestamp: 2026-01-12T10:58:40.230Z
Learning: In lightspeed-core/lightspeed-stack, for Python files under src/models, when a user claims a fix is done but the issue persists, verify the current code state before accepting the fix. Steps: review the diff, fetch the latest changes, run relevant tests, reproduce the issue, search the codebase for lingering references to the original problem, confirm the fix is applied and not undone by subsequent commits, and validate with local checks to ensure the issue is resolved.
Applied to files:
src/models/api/requests/conversations.pysrc/models/api/requests/__init__.pysrc/models/common/__init__.pysrc/models/common/feedback_category.pysrc/models/context.pysrc/models/api/requests/catalog.pysrc/models/common/query.pysrc/models/common/health.pysrc/models/api/requests/prompts.pysrc/models/api/requests/feedback.pysrc/models/api/requests/responses_openai.pysrc/models/api/requests/mcp_servers.pysrc/models/api/requests/vector_stores.pysrc/models/api/requests/query.py
📚 Learning: 2026-02-25T07:46:33.545Z
Learnt from: asimurka
Repo: lightspeed-core/lightspeed-stack PR: 1211
File: src/models/responses.py:8-16
Timestamp: 2026-02-25T07:46:33.545Z
Learning: In the Python codebase, requests.py should use OpenAIResponseInputTool as Tool while responses.py uses OpenAIResponseTool as Tool. This difference is intentional due to differing schemas for input vs output tools in llama-stack-api. Apply this distinction consistently to other models under src/models (e.g., ensure request-related tools use the InputTool variant and response-related tools use the ResponseTool variant). If adding new tools, choose the corresponding InputTool or Tool class based on whether the tool represents input or output, and document the rationale in code comments.
Applied to files:
src/models/api/requests/conversations.pysrc/models/api/requests/__init__.pysrc/models/common/__init__.pysrc/models/common/feedback_category.pysrc/models/context.pysrc/models/api/requests/catalog.pysrc/models/common/query.pysrc/models/common/health.pysrc/models/api/requests/prompts.pysrc/models/api/requests/feedback.pysrc/models/api/requests/responses_openai.pysrc/models/api/requests/mcp_servers.pysrc/models/api/requests/vector_stores.pysrc/models/api/requests/query.py
📚 Learning: 2026-03-02T15:57:10.746Z
Learnt from: asamal4
Repo: lightspeed-core/lightspeed-stack PR: 1250
File: AGENTS.md:11-12
Timestamp: 2026-03-02T15:57:10.746Z
Learning: In AGENTS.md, Linting Tools section should list actual tool names (e.g., mypy, pylint, pyright, ruff, black) instead of make target names (e.g., check-types). Ensure the section documents the individual tools that perform linting, not the Makefile targets that invoke them, so readers understand which tools are used and how to configure/run them directly.
Applied to files:
AGENTS.md
📚 Learning: 2026-04-06T20:18:07.852Z
Learnt from: major
Repo: lightspeed-core/lightspeed-stack PR: 1463
File: src/app/endpoints/rlsapi_v1.py:266-271
Timestamp: 2026-04-06T20:18:07.852Z
Learning: In the lightspeed-stack codebase, within `src/app/endpoints/` inference/MCP endpoints, treat `tools: Optional[list[Any]]` in MCP tool definitions as an intentional, consistent typing pattern (used across `query`, `responses`, `streaming_query`, `rlsapi_v1`). Do not raise or suggest this as a typing issue during code review; changing it in isolation could break endpoint typing consistency across the codebase.
Applied to files:
src/app/endpoints/mcp_servers.pysrc/app/endpoints/vector_stores.pysrc/app/endpoints/a2a.pysrc/app/endpoints/feedback.pysrc/app/endpoints/stream_interrupt.pysrc/app/endpoints/prompts.pysrc/app/endpoints/query.pysrc/app/endpoints/conversations_v2.pysrc/app/endpoints/responses.pysrc/app/endpoints/conversations_v1.pysrc/app/endpoints/health.pysrc/app/endpoints/models.pysrc/app/endpoints/streaming_query.py
📚 Learning: 2026-01-14T09:37:51.612Z
Learnt from: asimurka
Repo: lightspeed-core/lightspeed-stack PR: 988
File: src/app/endpoints/query.py:319-339
Timestamp: 2026-01-14T09:37:51.612Z
Learning: In the lightspeed-stack repository, when provider_id == "azure", the Azure provider with provider_type "remote::azure" is guaranteed to be present in the providers list. Therefore, avoid defensive StopIteration handling for next() when locating the Azure provider in providers within src/app/endpoints/query.py. This change applies specifically to this file (or nearby provider lookup code) and relies on the invariant that the Azure provider exists; if the invariant could be violated, keep the existing StopIteration handling.
Applied to files:
src/app/endpoints/query.py
📚 Learning: 2026-02-23T14:56:59.186Z
Learnt from: asimurka
Repo: lightspeed-core/lightspeed-stack PR: 1198
File: src/utils/responses.py:184-192
Timestamp: 2026-02-23T14:56:59.186Z
Learning: In the lightspeed-stack codebase (lightspeed-core/lightspeed-stack), do not enforce de-duplication of duplicate client.models.list() calls in model selection flows (e.g., in src/utils/responses.py prepare_responses_params). These calls are considered relatively cheap and removing duplicates could add unnecessary complexity to the flow. Apply this guideline specifically to this file/context unless similar performance characteristics and design decisions are documented elsewhere.
Applied to files:
src/utils/responses.py
🪛 GitHub Actions: OpenAPI (Spectral)
docs/openapi.json
[error] 1-1: CI check failed: docs/openapi.json is out of date. Regenerate with: uv run scripts/generate_openapi_schema.py docs/openapi.json (diff -u /tmp/openapi-generated.json vs docs/openapi.json failed).
🪛 markdownlint-cli2 (0.22.1)
src/models/README.md
[warning] 15-15: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
[warning] 18-18: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
[warning] 21-21: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
[warning] 24-24: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
🔇 Additional comments (62)
tests/unit/models/requests/__init__.py (1)
1-1: Docstring update is consistent with the new model layout.Looks good and aligned with the refactor from
requests.pytomodels.api.requests.tests/unit/models/requests/README.md (1)
4-4: README entry correctly reflects the new request-model module path.This keeps docs in sync with the refactor.
AGENTS.md (1)
34-38: LGTM — project structure accurately reflects the new layout.The updated tree correctly documents the split into
api/requests/,api/responses/, andcommon/.src/models/common/feedback_category.py (1)
1-19: LGTM — clean, well-documented enum.Enum values are consistent (all lowercase), class and module docstrings are present, and the inline comments provide useful context for each category.
src/models/api/requests/conversations.py (1)
1-22: LGTM — well-structured request model with proper validation constraints.Field constraints (
min_length=1,max_length=1000) andextra="forbid"are all correctly applied.src/models/api/requests/prompts.py (1)
1-86: LGTM — both request models are well-structured with appropriate validation.
min_length,gt=0, andextra="forbid"constraints are correctly applied; OpenAPI examples are provided viajson_schema_extra.src/models/common/query.py (1)
89-109: LGTM — backward-compatibility validator is correct and well-documented.The three-branch guard (
None/non-dict→ pass-through,"filters"/"mode"present → new format, otherwise → legacy wrap) correctly handles all input shapes, and the deprecation warning message is precise and actionable.src/utils/vector_search.py (1)
21-21: Import relocation looks correct.The move to
from models.common.query import SolrVectorSearchRequestis consistent with the model split and keeps internal imports absolute.src/models/common/__init__.py (1)
1-51:models.commonexport surface is coherent after the split.The new exports and
__all__updates are consistent with the refactor and keep shared types centrally accessible.src/app/endpoints/health.py (1)
29-29: Shared health model import is correctly wired.Using
HealthStatus/ProviderHealthStatusfrommodels.commonmatches the new shared-model structure.tests/unit/app/endpoints/test_health.py (1)
17-17: Unit test import update is correct.Switching
HealthStatustomodels.commonkeeps tests aligned with the endpoint refactor.tests/integration/endpoints/test_health_integration.py (1)
17-17: Integration test import aligns with shared enum location.src/models/api/requests/responses_openai.py (1)
109-141: Body-size guard placement and error messaging look good.Pre-validation size checks are clear and enforce the configured cap consistently.
tests/unit/app/endpoints/test_models.py (1)
14-14: Import path migration is correct.This keeps the test aligned with the new request-model package surface.
tests/unit/models/requests/test_responses_request.py (1)
9-9: Updated import is consistent with the new model layout.tests/integration/endpoints/test_conversations_v2_integration.py (1)
22-22: Import update is correct for the refactored request modules.src/app/endpoints/stream_interrupt.py (1)
10-10: Request import swap is clean and consistent with endpoint typing.tests/unit/utils/test_responses.py (1)
62-62:QueryRequestimport path update looks good.tests/integration/endpoints/test_conversations_v1_integration.py (1)
23-23: Import migration is correct and keeps v1 tests aligned with the new package exports.tests/unit/app/endpoints/test_mcp_servers.py (1)
17-17: Import path update is correct for the refactored request model structure.tests/unit/app/endpoints/test_responses.py (1)
29-29: Import migration is correct and aligned with the request-model split.This update keeps the test wired to the new API request model package without changing test behavior.
tests/unit/utils/test_transcripts.py (1)
8-8: Good import-path update forQueryRequest.This keeps transcripts tests aligned with the new request model module layout.
tests/unit/app/endpoints/test_feedback.py (1)
21-21: Import relocation looks right for feedback request models.Both request types now correctly target the new API request package.
tests/unit/models/requests/test_vector_store_requests.py (1)
6-6: Vector store request imports are updated correctly.The tests now reference the new canonical request-model namespace.
tests/unit/app/endpoints/test_vector_stores.py (1)
26-30: This consolidated import move is correct.
VectorStoreCreateRequest,VectorStoreFileCreateRequest, andVectorStoreUpdateRequestare all sourced from the new API requests package as expected.tests/integration/endpoints/test_streaming_query_byok_integration.py (1)
19-19: Import update is aligned for BYOK streaming integration tests.Using
models.api.requests.QueryRequesthere is consistent with the rest of the refactor.tests/unit/app/endpoints/test_responses_splunk.py (1)
25-25:ResponsesRequestimport path migration is correct.This keeps the Splunk response tests synchronized with the new API request model organization.
src/utils/responses.py (1)
94-94: LGTM!Import correctly updated to the new
models.api.requestsnamespace, consistent with the PR-wide migration.tests/unit/models/requests/test_feedback_request.py (1)
6-6: LGTM!Import correctly migrated from the legacy
models.requeststomodels.api.requests.tests/unit/models/requests/test_attachment.py (1)
3-3: LGTM!
Attachmentcorrectly re-pointed to its new canonical home inmodels.common.query.tests/integration/endpoints/test_stream_interrupt_integration.py (1)
10-10: LGTM!Import correctly updated to
models.api.requests; test logic and lifecycle assertions are unaffected.tests/integration/endpoints/test_model_list.py (1)
15-15: LGTM!
ModelFilternow correctly imported frommodels.api.requests; all test scenarios are unaffected.tests/unit/app/endpoints/test_conversations_v2.py (1)
23-23: LGTM!
ConversationUpdateRequestcorrectly re-pointed tomodels.api.requests; no test logic affected.tests/unit/models/requests/test_feedback_status_update_request.py (1)
3-3: LGTM!Import correctly migrated to
models.api.requests.tests/unit/utils/test_prompts.py (1)
9-10: LGTM!
QueryRequestcorrectly sourced from the newmodels.api.requestsnamespace; all prompt-related test cases are unaffected.src/app/endpoints/feedback.py (1)
16-16: Import migration is correct and consistent.Using
models.api.requestshere aligns with the request-model split and keeps endpoint typing unchanged.tests/unit/app/endpoints/test_prompts.py (1)
20-20: Test import update looks good.The request model source now matches the new API request package used by production code.
src/app/endpoints/conversations_v1.py (1)
19-19: Good import source update for conversation request model.This keeps the endpoint aligned with the new request model module layout.
tests/integration/endpoints/test_query_byok_integration.py (1)
18-18: Integration test import migration is consistent.
QueryRequestnow comes from the same package structure used across the refactor.tests/unit/app/endpoints/test_stream_interrupt.py (1)
10-10: Looks good — request model import was migrated correctly.No functional behavior is affected by this test-side update.
src/app/endpoints/vector_stores.py (1)
25-29: Request model imports are correctly re-pointed.All vector store request types are now sourced from the new API request package as expected.
tests/unit/app/endpoints/test_conversations.py (1)
22-22: Conversation update request import update is correct.The unit tests now reference the same request-model package as the endpoint code.
src/app/endpoints/a2a.py (1)
47-47: A2A request model import migration is clean.
QueryRequestis now sourced from the consolidated API request package consistently.tests/unit/utils/test_vector_search.py (1)
9-9: Import path update looks correct.This aligns with the request/common model split and is consistent with usage in this test module.
tests/integration/endpoints/test_query_integration.py (1)
21-23: Request/attachment imports are correctly migrated.The new import sources match the refactored model package boundaries without changing test behavior.
tests/integration/endpoints/test_streaming_query_integration.py (1)
15-16: Streaming integration imports are aligned with the new model layout.No functional risk here; the updated modules are the correct targets for these types.
src/app/endpoints/query.py (1)
27-27: Endpoint request type import is correctly updated.This keeps
/querybound to the new API request-model surface.src/app/endpoints/mcp_servers.py (1)
14-14: MCP registration request import migration is correct.The handler remains correctly typed against the new API request module.
tests/unit/app/endpoints/test_streaming_query.py (1)
67-70: Unit test imports are migrated consistently.The updated paths reflect the new request/common model separation without altering behavior.
tests/unit/utils/test_query.py (1)
18-25:utils.queryunit tests now reference the right model modules.Good alignment with the new API/common request model structure.
src/app/endpoints/models.py (1)
15-15:ModelFilterimport update is appropriate.The
/modelsendpoint remains correctly typed against the refactored catalog request model.src/utils/transcripts.py (1)
19-19: LGTM – import path correctly updated.tests/unit/app/endpoints/test_query.py (1)
15-18: LGTM – imports correctly updated to new module locations.src/models/context.py (1)
7-7: LGTM –QueryRequestcorrectly sourced from new location.tests/unit/models/requests/test_query_request.py (1)
5-6: LGTM – imports correctly updated.src/app/endpoints/streaming_query.py (1)
63-63: LGTM –QueryRequestimport correctly updated tomodels.api.requests.src/utils/shields.py (1)
21-21: LGTM –QueryRequestimport correctly updated.src/app/endpoints/conversations_v2.py (1)
11-11: LGTM –ConversationUpdateRequestimport correctly updated.src/app/endpoints/prompts.py (1)
16-16: LGTM –PromptCreateRequestandPromptUpdateRequestimports correctly updated.src/app/endpoints/responses.py (1)
41-41: LGTM — import re-pointed correctly.
ResponsesRequestis properly exported frommodels.api.requests.__init__(confirmed insrc/models/api/requests/__init__.py, line 9 /__all__line 27) and defined insrc/models/api/requests/responses_openai.py.src/utils/query.py (1)
28-28: LGTM — both imports re-pointed correctly.
QueryRequestis re-exported frommodels.api.requestsandAttachmentis defined inmodels.common.query; both usages inprepare_input,store_query_results, andvalidate_attachments_metadataare unaffected.Also applies to: 39-39
src/models/api/requests/__init__.py (1)
1-32: LGTM — aggregation module is well-structured.All submodule imports resolve to the new domain-organized files,
__all__is exhaustive, and theFeedbackCategoryconvenience re-export keeps callers from needing a second import for constructingFeedbackRequest.
asimurka
force-pushed
the
move_error_responses
branch
from
066674f to
2c45fa3
Compare
| HEALTHY = "healthy" | ||
| UNKNOWN = "unknown" | ||
|
|
||
|
|
There was a problem hiding this comment.
Replaced out of endpoint to src/models/common/health.py
asimurka
force-pushed
the
move_error_responses
branch
from
2c45fa3 to
651c12b
Compare
There was a problem hiding this comment.
Actionable comments posted: 4
♻️ Duplicate comments (8)
src/models/README.md (1)
15-25:⚠️ Potential issue | 🟡 Minor | ⚡ Quick winMD022: Headings still missing blank lines below them.
The four new headings (lines 15, 18, 21, 24) are each immediately followed by description text without a blank line, which still triggers
markdownlintMD022. This was flagged in a previous review and marked addressed, but the current code does not reflect the fix.📄 Proposed fix
-## [api/](api/) -Models for API request and response bodies. +## [api/](api/) + +Models for API request and response bodies. -## [common/](common/) -Shared models and types used across endpoints. +## [common/](common/) + +Shared models and types used across endpoints. -## [database/](database/) -SQLAlchemy ORM models. +## [database/](database/) + +SQLAlchemy ORM models. -## [rlsapi/](rlsapi/) -Models for the rlsapi v1 API surface. +## [rlsapi/](rlsapi/) + +Models for the rlsapi v1 API surface.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/README.md` around lines 15 - 25, The four level-2 headings "## [api/](api/)", "## [common/](common/)", "## [database/](database/)", and "## [rlsapi/](rlsapi/)" in README.md lack the required blank line after each heading which triggers markdownlint MD022; fix this by inserting a single empty line immediately after each of those heading lines so each heading is followed by a blank line before its descriptive text, then run markdownlint to verify the MD022 warning is resolved.src/models/api/requests/catalog.py (1)
12-12:⚠️ Potential issue | 🟡 Minor | ⚡ Quick winMisleading docstring: "Required model type" should be "Optional model type filter".
The
Attributessection still says "Required model type" while the field isOptional[str]defaulting toNone. This was flagged in a prior review and remains unresolved.📄 Proposed fix
- model_type: Required model type, such as 'llm', 'embeddings' etc. + model_type: Optional filter to narrow results to models of this type, such as 'llm', 'embeddings' etc.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/api/requests/catalog.py` at line 12, Update the misleading docstring for the model_type attribute in src/models/api/requests/catalog.py: change "Required model type" to "Optional model type filter" and clarify it is Optional[str] defaulting to None (e.g., "Optional model type filter, such as 'llm' or 'embeddings'; defaults to None"). Ensure the description next to the model_type field and any Attributes section consistently reflect that it's an optional filter rather than required.src/models/api/requests/vector_stores.py (1)
3-3:⚠️ Potential issue | 🟡 Minor | ⚡ Quick win
Selfmust be imported fromtyping_extensions, nottyping.The coding guidelines explicitly require
typing_extensions.Self, and the project has Ruff rule UP035 enabled. This was flagged in a prior review and remains unresolved.📄 Proposed fix
-from typing import Any, Optional, Self +from typing import Any, Optional +from typing_extensions import Self🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/api/requests/vector_stores.py` at line 3, The import is using Self from the stdlib typing which violates the UP035 rule; change the import statement so that Self is imported from typing_extensions instead of typing (e.g., replace "from typing import Any, Optional, Self" with importing Any and Optional from typing and Self from typing_extensions), ensuring the symbol Self used in this module (e.g., any class or function type hints referencing Self) now references typing_extensions.Self.src/models/common/health.py (1)
9-16:⚠️ Potential issue | 🟡 Minor | ⚡ Quick win
ERROR = "Error"still has inconsistent casing.Every other member (
"ok","not_implemented","healthy","unknown") is lowercase. The capital"E"will silently mismatch any code that compares.valueto a lowercase string literal.🐛 Proposed fix
- ERROR = "Error" + ERROR = "error"🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/common/health.py` around lines 9 - 16, The HealthStatus enum has inconsistent casing: the member ERROR in class HealthStatus currently uses "Error" which should be lowercase; update HealthStatus.ERROR to use "error" to match the other members and avoid silent mismatches when comparing HealthStatus.value to lowercase strings (check usages of HealthStatus.ERROR if any to ensure comparisons remain valid).tests/unit/models/requests/test_query_request.py (1)
95-109:⚠️ Potential issue | 🟡 Minor | ⚡ Quick win
test_validate_media_typeis still missing the rejection andapplication/jsonbranches.Only
text/plainacceptance is covered. Thevalidate_media_typevalidator raisesValueErrorfor unsupported types, andapplication/jsonis also a valid value — both branches remain untested.🛡️ Proposed additions
+ def test_validate_media_type_json(self) -> None: + """Test that application/json is also an accepted media_type.""" + qr = QueryRequest( + query="Tell me about Kubernetes", + media_type="application/json", + ) # pyright: ignore[reportCallIssue] + assert qr.media_type == "application/json" + + def test_validate_media_type_invalid_raises(self) -> None: + """Test that an unsupported media_type raises ValueError.""" + with pytest.raises(ValueError, match="media_type must be either"): + QueryRequest( + query="Tell me about Kubernetes", + media_type="text/html", + ) # pyright: ignore[reportCallIssue]🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@tests/unit/models/requests/test_query_request.py` around lines 95 - 109, The test test_validate_media_type only checks that QueryRequest stores "text/plain"; add two more assertions: one that QueryRequest(media_type="application/json") is accepted and stores media_type "application/json", and one that an unsupported media_type (e.g., "image/png") raises ValueError from the validate_media_type validator when constructing QueryRequest; use pytest.raises(ValueError) to assert the rejection and direct attribute checks for the successful application/json case, referencing the QueryRequest class and its validate_media_type validator.src/models/api/requests/feedback.py (1)
170-179:⚠️ Potential issue | 🟡 Minor | ⚡ Quick winWhitespace-only
user_feedbackstill bypasses the "at least one feedback signal" guard.
" "is neitherNonenor"", so(self.user_feedback is None or self.user_feedback == "")evaluates toFalse, allowing a blank-whitespace-only string to satisfy the presence check.Proposed fix
if ( self.sentiment is None - and (self.user_feedback is None or self.user_feedback == "") + and not (self.user_feedback or "").strip() and self.categories is None ):🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/api/requests/feedback.py` around lines 170 - 179, The presence check currently treats whitespace-only user_feedback as valid; update the guard in the method that returns self to treat strings containing only whitespace as empty by checking user_feedback.strip() (e.g., consider user_feedback is None or user_feedback.strip() == "" or use a truthy check like not (self.user_feedback and self.user_feedback.strip())); ensure the conditional that combines sentiment, user_feedback, and categories uses this trimmed check so whitespace-only user_feedback no longer bypasses the validation for attributes sentiment, user_feedback, and categories.src/models/api/requests/responses_openai.py (1)
157-170:⚠️ Potential issue | 🟡 Minor | ⚡ Quick win
conversationempty-string bypass in mutual-exclusivity check and SUID validator still unresolved.Both conditions treat
""as "not provided" because they use truthiness rather thanis not None, silently skipping validation:
- Line 157:
if self.conversation and self.previous_response_id— a client sendingconversation=""alongsideprevious_response_idpasses the mutual-exclusivity guard.- Line 168:
if value and not suid.check_suid(value)—conversation=""skips SUID validation entirely.Proposed fix
- if self.conversation and self.previous_response_id: + if self.conversation is not None and self.previous_response_id is not None: raise ValueError( "`conversation` and `previous_response_id` are mutually exclusive. " "Only one can be provided at a time." )- if value and not suid.check_suid(value): + if value is not None and not suid.check_suid(value): raise ValueError(f"Improper conversation ID '{value}'")🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/api/requests/responses_openai.py` around lines 157 - 170, The mutual-exclusivity and SUID validation incorrectly use truthiness so an empty string ("") bypasses checks; update the guard in the method that checks `self.conversation` and `self.previous_response_id` to treat an empty string as provided (e.g., use explicit checks like `self.conversation is not None and self.conversation != ""`), and update the `check_suid` field validator (the `@field_validator` method named `check_suid`) to only skip validation for None but validate and reject empty-string and non-SUID values (e.g., `if value is not None and value != "" and not suid.check_suid(value): raise ValueError(...)`).src/models/common/query.py (1)
38-59: 🧹 Nitpick | 🔵 Trivial | 💤 Low value
Attachment.model_configstill uses a raw dict whileSolrVectorSearchRequestin the same file usesConfigDict.This inconsistency was raised in a previous review and remains unresolved.
♻️ Proposed fix
- model_config = { - "extra": "forbid", - "json_schema_extra": { - "examples": [ - ... - ] - }, - } + model_config = ConfigDict( + extra="forbid", + json_schema_extra={ + "examples": [ + ... + ] + }, + )🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/models/common/query.py` around lines 38 - 59, Attachment.model_config is using a plain dict while SolrVectorSearchRequest uses pydantic's ConfigDict; replace the raw dict with a typed ConfigDict for Attachment.model_config, import ConfigDict from pydantic (or reuse the existing import pattern used by SolrVectorSearchRequest), and assign model_config as a ClassVar[ConfigDict] containing the same keys ("extra": "forbid", "json_schema_extra": {...}) to match the other model's style and ensure consistency.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@src/app/endpoints/models.py`:
- Line 15: Replace the direct submodule import with the package-level re-export:
change the import of ModelFilter in endpoints/models.py from
models.api.requests.catalog to using the package import (ModelFilter from
models.api.requests) so it matches tests and other modules; update the import
statement that references ModelFilter accordingly to use the package-level
symbol exported by models.api.requests.
In `@src/models/api/requests/mcp_servers.py`:
- Around line 145-154: The code validates header values by using stripped =
header_value.strip() but returns the original value dict unchanged, so inputs
like " client " remain unnormalized; update the loop in the validator to store
the normalized value back into the dict (e.g., value[header_name] = stripped) or
build and return a new dict of stripped values after checking allowed (use the
existing symbols header_name, header_value, stripped, value, allowed) so the
function returns normalized header strings rather than the original untrimmed
ones.
In `@src/models/api/requests/query.py`:
- Line 3: Replace the use of typing.Self with typing_extensions.Self in the
import statement so validators in Pydantic models use the portable,
Pydantic-recommended Self; update the line that currently imports "Self" (e.g.,
"from typing import Optional, Self") to import Self from typing_extensions
instead, ensuring all references to Self in model methods/validators (such as
any Pydantic model classes or validator signatures in this file) remain
unchanged.
- Around line 81-91: The fields no_tools and generate_topic_summary are typed as
Optional[bool] which allows callers to send null and makes downstream truthiness
checks unreliable; change their annotations from Optional[bool] to plain bool
and keep the existing default values (no_tools=False,
generate_topic_summary=True) so Pydantic will reject nulls and preserve intended
boolean semantics for the no_tools and generate_topic_summary fields defined
with Field(...).
---
Duplicate comments:
In `@src/models/api/requests/catalog.py`:
- Line 12: Update the misleading docstring for the model_type attribute in
src/models/api/requests/catalog.py: change "Required model type" to "Optional
model type filter" and clarify it is Optional[str] defaulting to None (e.g.,
"Optional model type filter, such as 'llm' or 'embeddings'; defaults to None").
Ensure the description next to the model_type field and any Attributes section
consistently reflect that it's an optional filter rather than required.
In `@src/models/api/requests/feedback.py`:
- Around line 170-179: The presence check currently treats whitespace-only
user_feedback as valid; update the guard in the method that returns self to
treat strings containing only whitespace as empty by checking
user_feedback.strip() (e.g., consider user_feedback is None or
user_feedback.strip() == "" or use a truthy check like not (self.user_feedback
and self.user_feedback.strip())); ensure the conditional that combines
sentiment, user_feedback, and categories uses this trimmed check so
whitespace-only user_feedback no longer bypasses the validation for attributes
sentiment, user_feedback, and categories.
In `@src/models/api/requests/responses_openai.py`:
- Around line 157-170: The mutual-exclusivity and SUID validation incorrectly
use truthiness so an empty string ("") bypasses checks; update the guard in the
method that checks `self.conversation` and `self.previous_response_id` to treat
an empty string as provided (e.g., use explicit checks like `self.conversation
is not None and self.conversation != ""`), and update the `check_suid` field
validator (the `@field_validator` method named `check_suid`) to only skip
validation for None but validate and reject empty-string and non-SUID values
(e.g., `if value is not None and value != "" and not suid.check_suid(value):
raise ValueError(...)`).
In `@src/models/api/requests/vector_stores.py`:
- Line 3: The import is using Self from the stdlib typing which violates the
UP035 rule; change the import statement so that Self is imported from
typing_extensions instead of typing (e.g., replace "from typing import Any,
Optional, Self" with importing Any and Optional from typing and Self from
typing_extensions), ensuring the symbol Self used in this module (e.g., any
class or function type hints referencing Self) now references
typing_extensions.Self.
In `@src/models/common/health.py`:
- Around line 9-16: The HealthStatus enum has inconsistent casing: the member
ERROR in class HealthStatus currently uses "Error" which should be lowercase;
update HealthStatus.ERROR to use "error" to match the other members and avoid
silent mismatches when comparing HealthStatus.value to lowercase strings (check
usages of HealthStatus.ERROR if any to ensure comparisons remain valid).
In `@src/models/common/query.py`:
- Around line 38-59: Attachment.model_config is using a plain dict while
SolrVectorSearchRequest uses pydantic's ConfigDict; replace the raw dict with a
typed ConfigDict for Attachment.model_config, import ConfigDict from pydantic
(or reuse the existing import pattern used by SolrVectorSearchRequest), and
assign model_config as a ClassVar[ConfigDict] containing the same keys ("extra":
"forbid", "json_schema_extra": {...}) to match the other model's style and
ensure consistency.
In `@src/models/README.md`:
- Around line 15-25: The four level-2 headings "## [api/](api/)", "##
[common/](common/)", "## [database/](database/)", and "## [rlsapi/](rlsapi/)" in
README.md lack the required blank line after each heading which triggers
markdownlint MD022; fix this by inserting a single empty line immediately after
each of those heading lines so each heading is followed by a blank line before
its descriptive text, then run markdownlint to verify the MD022 warning is
resolved.
In `@tests/unit/models/requests/test_query_request.py`:
- Around line 95-109: The test test_validate_media_type only checks that
QueryRequest stores "text/plain"; add two more assertions: one that
QueryRequest(media_type="application/json") is accepted and stores media_type
"application/json", and one that an unsupported media_type (e.g., "image/png")
raises ValueError from the validate_media_type validator when constructing
QueryRequest; use pytest.raises(ValueError) to assert the rejection and direct
attribute checks for the successful application/json case, referencing the
QueryRequest class and its validate_media_type validator.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro
Run ID: 299316b8-10d0-4861-95ef-03749c488eec
📒 Files selected for processing (71)
AGENTS.mddocs/openapi.jsonsrc/app/endpoints/a2a.pysrc/app/endpoints/conversations_v1.pysrc/app/endpoints/conversations_v2.pysrc/app/endpoints/feedback.pysrc/app/endpoints/health.pysrc/app/endpoints/mcp_servers.pysrc/app/endpoints/models.pysrc/app/endpoints/prompts.pysrc/app/endpoints/query.pysrc/app/endpoints/responses.pysrc/app/endpoints/stream_interrupt.pysrc/app/endpoints/streaming_query.pysrc/app/endpoints/vector_stores.pysrc/models/README.mdsrc/models/api/requests/__init__.pysrc/models/api/requests/catalog.pysrc/models/api/requests/conversations.pysrc/models/api/requests/feedback.pysrc/models/api/requests/mcp_servers.pysrc/models/api/requests/prompts.pysrc/models/api/requests/query.pysrc/models/api/requests/responses_openai.pysrc/models/api/requests/vector_stores.pysrc/models/common/__init__.pysrc/models/common/feedback.pysrc/models/common/health.pysrc/models/common/query.pysrc/models/context.pysrc/models/requests.pysrc/utils/query.pysrc/utils/responses.pysrc/utils/shields.pysrc/utils/transcripts.pysrc/utils/vector_search.pytests/integration/endpoints/test_conversations_v1_integration.pytests/integration/endpoints/test_conversations_v2_integration.pytests/integration/endpoints/test_health_integration.pytests/integration/endpoints/test_model_list.pytests/integration/endpoints/test_query_byok_integration.pytests/integration/endpoints/test_query_integration.pytests/integration/endpoints/test_stream_interrupt_integration.pytests/integration/endpoints/test_streaming_query_byok_integration.pytests/integration/endpoints/test_streaming_query_integration.pytests/unit/app/endpoints/test_conversations.pytests/unit/app/endpoints/test_conversations_v2.pytests/unit/app/endpoints/test_feedback.pytests/unit/app/endpoints/test_health.pytests/unit/app/endpoints/test_mcp_servers.pytests/unit/app/endpoints/test_models.pytests/unit/app/endpoints/test_prompts.pytests/unit/app/endpoints/test_query.pytests/unit/app/endpoints/test_responses.pytests/unit/app/endpoints/test_responses_splunk.pytests/unit/app/endpoints/test_stream_interrupt.pytests/unit/app/endpoints/test_streaming_query.pytests/unit/app/endpoints/test_vector_stores.pytests/unit/models/requests/README.mdtests/unit/models/requests/__init__.pytests/unit/models/requests/test_attachment.pytests/unit/models/requests/test_feedback_request.pytests/unit/models/requests/test_feedback_status_update_request.pytests/unit/models/requests/test_query_request.pytests/unit/models/requests/test_responses_request.pytests/unit/models/requests/test_vector_store_requests.pytests/unit/utils/test_prompts.pytests/unit/utils/test_query.pytests/unit/utils/test_responses.pytests/unit/utils/test_transcripts.pytests/unit/utils/test_vector_search.py
💤 Files with no reviewable changes (1)
- src/models/requests.py
📜 Review details
🧰 Additional context used
📓 Path-based instructions (5)
src/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Use absolute imports for internal modules:
from authentication import get_auth_dependencyUse
from llama_stack_client import AsyncLlamaStackClientfor Llama Stack integrationAll modules must start with descriptive docstrings explaining their purpose
Use
logger = get_logger(__name__)fromlog.pyfor module loggingUse
Final[type]as type hint for all constantsDefine type aliases at module level for clarity
Use
@field_validatorand@model_validatorfor custom validation in Pydantic modelsAll functions must have complete docstrings with brief descriptions
Use complete type annotations for all function parameters and return types
Use
typing_extensions.Selffor model validators in Pydantic modelsUse modern union syntax
str | intinstead ofUnion[str, int]for type annotationsUse
Optional[Type]for optional parameters in type annotationsUse snake_case with descriptive, action-oriented function names (get_, validate_, check_ patterns)
Avoid in-place parameter modification anti-patterns; return new data structures instead
Use
async deffor functions that perform I/O operations and external API callsHandle
APIConnectionErrorfrom Llama Stack in functions that interact with Llama StackUse standard log levels with clear purposes: debug (diagnostic), info (execution), warning (unexpected), error (serious problems)
All classes must have descriptive docstrings explaining their purpose
Use PascalCase for class names with descriptive names and standard suffixes (Configuration, Error/Exception, Resolver, Interface)
Pydantic configuration models must extend
ConfigurationBase, data models must extendBaseModelUse ABC (Abstract Base Class) for interfaces with
@abstractmethoddecoratorsUse complete type annotations for all class attributes; avoid using
AnytypeFollow Google Python docstring conventions (https://google.github.io/styleguide/pyguide.html) with Parameters, Returns, Raises, and Attributes sections
Files:
src/app/endpoints/feedback.pysrc/utils/responses.pysrc/models/common/feedback.pysrc/models/api/requests/__init__.pysrc/utils/vector_search.pysrc/models/common/health.pysrc/app/endpoints/stream_interrupt.pysrc/models/api/requests/vector_stores.pysrc/app/endpoints/a2a.pysrc/app/endpoints/streaming_query.pysrc/app/endpoints/vector_stores.pysrc/models/api/requests/mcp_servers.pysrc/models/common/__init__.pysrc/app/endpoints/query.pysrc/app/endpoints/models.pysrc/models/api/requests/prompts.pysrc/models/common/query.pysrc/models/api/requests/conversations.pysrc/models/api/requests/catalog.pysrc/app/endpoints/conversations_v1.pysrc/app/endpoints/responses.pysrc/utils/shields.pysrc/utils/transcripts.pysrc/app/endpoints/mcp_servers.pysrc/models/api/requests/query.pysrc/utils/query.pysrc/app/endpoints/prompts.pysrc/models/context.pysrc/models/api/requests/feedback.pysrc/app/endpoints/conversations_v2.pysrc/models/api/requests/responses_openai.pysrc/app/endpoints/health.py
src/app/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Use FastAPI dependencies:
from fastapi import APIRouter, HTTPException, Request, status, DependsUse FastAPI
HTTPExceptionwith appropriate status codes for API endpoint error handling
Files:
src/app/endpoints/feedback.pysrc/app/endpoints/stream_interrupt.pysrc/app/endpoints/a2a.pysrc/app/endpoints/streaming_query.pysrc/app/endpoints/vector_stores.pysrc/app/endpoints/query.pysrc/app/endpoints/models.pysrc/app/endpoints/conversations_v1.pysrc/app/endpoints/responses.pysrc/app/endpoints/mcp_servers.pysrc/app/endpoints/prompts.pysrc/app/endpoints/conversations_v2.pysrc/app/endpoints/health.py
tests/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Use pytest for all unit and integration tests; do not use unittest
Use
pytest-mockwith AsyncMock objects for mocking in unit testsUse
pytest.mark.asynciomarker for async test functions
Files:
tests/unit/models/requests/test_feedback_request.pytests/unit/models/requests/test_attachment.pytests/integration/endpoints/test_stream_interrupt_integration.pytests/unit/models/requests/__init__.pytests/unit/utils/test_prompts.pytests/unit/app/endpoints/test_feedback.pytests/unit/app/endpoints/test_conversations_v2.pytests/unit/app/endpoints/test_mcp_servers.pytests/unit/app/endpoints/test_responses.pytests/unit/utils/test_responses.pytests/unit/app/endpoints/test_streaming_query.pytests/integration/endpoints/test_conversations_v2_integration.pytests/integration/endpoints/test_conversations_v1_integration.pytests/unit/app/endpoints/test_prompts.pytests/unit/models/requests/test_feedback_status_update_request.pytests/unit/models/requests/test_query_request.pytests/integration/endpoints/test_streaming_query_byok_integration.pytests/unit/models/requests/test_vector_store_requests.pytests/unit/utils/test_transcripts.pytests/unit/utils/test_query.pytests/unit/app/endpoints/test_models.pytests/unit/app/endpoints/test_query.pytests/integration/endpoints/test_health_integration.pytests/unit/utils/test_vector_search.pytests/unit/app/endpoints/test_conversations.pytests/unit/models/requests/test_responses_request.pytests/integration/endpoints/test_streaming_query_integration.pytests/unit/app/endpoints/test_stream_interrupt.pytests/integration/endpoints/test_query_byok_integration.pytests/integration/endpoints/test_model_list.pytests/unit/app/endpoints/test_vector_stores.pytests/integration/endpoints/test_query_integration.pytests/unit/app/endpoints/test_health.pytests/unit/app/endpoints/test_responses_splunk.py
src/**/__init__.py
📄 CodeRabbit inference engine (AGENTS.md)
Package
__init__.pyfiles must contain brief package descriptions
Files:
src/models/api/requests/__init__.pysrc/models/common/__init__.py
AGENTS.md
📄 CodeRabbit inference engine (CLAUDE.md)
Document agent implementations and their configurations in AGENTS.md
Files:
AGENTS.md
🧠 Learnings (7)
📓 Common learnings
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Use `uv run` prefix for all commands (format, verify, testing, etc.)
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Unit tests require 60% code coverage; integration tests require 10% coverage
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Run `uv run make format` to auto-format code with black and ruff before completion
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Run `uv run make verify` to run all linters (black, pylint, pyright, ruff, docstyle, check-types) before completion
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: PR titles MUST start with JIRA issue key prefix (LCORE-, RSPEED-, MGTM-, OLS-, RHDHPAI-, LEADS-)
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Never commit secrets/keys; use environment variables for sensitive data
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Use bandit for security issue detection
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Use `uv sync --group dev --group llslibdev` for setting up development dependencies
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Follow existing code patterns in the module you're modifying
Learnt from: CR
Repo: lightspeed-core/lightspeed-stack
Timestamp: 2026-05-05T12:39:25.013Z
Learning: Create unit tests covering new functionality before completion
📚 Learning: 2026-04-06T20:18:07.852Z
Learnt from: major
Repo: lightspeed-core/lightspeed-stack PR: 1463
File: src/app/endpoints/rlsapi_v1.py:266-271
Timestamp: 2026-04-06T20:18:07.852Z
Learning: In the lightspeed-stack codebase, within `src/app/endpoints/` inference/MCP endpoints, treat `tools: Optional[list[Any]]` in MCP tool definitions as an intentional, consistent typing pattern (used across `query`, `responses`, `streaming_query`, `rlsapi_v1`). Do not raise or suggest this as a typing issue during code review; changing it in isolation could break endpoint typing consistency across the codebase.
Applied to files:
src/app/endpoints/feedback.pysrc/app/endpoints/stream_interrupt.pysrc/app/endpoints/a2a.pysrc/app/endpoints/streaming_query.pysrc/app/endpoints/vector_stores.pysrc/app/endpoints/query.pysrc/app/endpoints/models.pysrc/app/endpoints/conversations_v1.pysrc/app/endpoints/responses.pysrc/app/endpoints/mcp_servers.pysrc/app/endpoints/prompts.pysrc/app/endpoints/conversations_v2.pysrc/app/endpoints/health.py
📚 Learning: 2026-02-23T14:56:59.186Z
Learnt from: asimurka
Repo: lightspeed-core/lightspeed-stack PR: 1198
File: src/utils/responses.py:184-192
Timestamp: 2026-02-23T14:56:59.186Z
Learning: In the lightspeed-stack codebase (lightspeed-core/lightspeed-stack), do not enforce de-duplication of duplicate client.models.list() calls in model selection flows (e.g., in src/utils/responses.py prepare_responses_params). These calls are considered relatively cheap and removing duplicates could add unnecessary complexity to the flow. Apply this guideline specifically to this file/context unless similar performance characteristics and design decisions are documented elsewhere.
Applied to files:
src/utils/responses.py
📚 Learning: 2026-01-12T10:58:40.230Z
Learnt from: blublinsky
Repo: lightspeed-core/lightspeed-stack PR: 972
File: src/models/config.py:459-513
Timestamp: 2026-01-12T10:58:40.230Z
Learning: In lightspeed-core/lightspeed-stack, for Python files under src/models, when a user claims a fix is done but the issue persists, verify the current code state before accepting the fix. Steps: review the diff, fetch the latest changes, run relevant tests, reproduce the issue, search the codebase for lingering references to the original problem, confirm the fix is applied and not undone by subsequent commits, and validate with local checks to ensure the issue is resolved.
Applied to files:
src/models/common/feedback.pysrc/models/api/requests/__init__.pysrc/models/common/health.pysrc/models/api/requests/vector_stores.pysrc/models/api/requests/mcp_servers.pysrc/models/common/__init__.pysrc/models/api/requests/prompts.pysrc/models/common/query.pysrc/models/api/requests/conversations.pysrc/models/api/requests/catalog.pysrc/models/api/requests/query.pysrc/models/context.pysrc/models/api/requests/feedback.pysrc/models/api/requests/responses_openai.py
📚 Learning: 2026-02-25T07:46:33.545Z
Learnt from: asimurka
Repo: lightspeed-core/lightspeed-stack PR: 1211
File: src/models/responses.py:8-16
Timestamp: 2026-02-25T07:46:33.545Z
Learning: In the Python codebase, requests.py should use OpenAIResponseInputTool as Tool while responses.py uses OpenAIResponseTool as Tool. This difference is intentional due to differing schemas for input vs output tools in llama-stack-api. Apply this distinction consistently to other models under src/models (e.g., ensure request-related tools use the InputTool variant and response-related tools use the ResponseTool variant). If adding new tools, choose the corresponding InputTool or Tool class based on whether the tool represents input or output, and document the rationale in code comments.
Applied to files:
src/models/common/feedback.pysrc/models/api/requests/__init__.pysrc/models/common/health.pysrc/models/api/requests/vector_stores.pysrc/models/api/requests/mcp_servers.pysrc/models/common/__init__.pysrc/models/api/requests/prompts.pysrc/models/common/query.pysrc/models/api/requests/conversations.pysrc/models/api/requests/catalog.pysrc/models/api/requests/query.pysrc/models/context.pysrc/models/api/requests/feedback.pysrc/models/api/requests/responses_openai.py
📚 Learning: 2026-03-02T15:57:10.746Z
Learnt from: asamal4
Repo: lightspeed-core/lightspeed-stack PR: 1250
File: AGENTS.md:11-12
Timestamp: 2026-03-02T15:57:10.746Z
Learning: In AGENTS.md, Linting Tools section should list actual tool names (e.g., mypy, pylint, pyright, ruff, black) instead of make target names (e.g., check-types). Ensure the section documents the individual tools that perform linting, not the Makefile targets that invoke them, so readers understand which tools are used and how to configure/run them directly.
Applied to files:
AGENTS.md
📚 Learning: 2026-01-14T09:37:51.612Z
Learnt from: asimurka
Repo: lightspeed-core/lightspeed-stack PR: 988
File: src/app/endpoints/query.py:319-339
Timestamp: 2026-01-14T09:37:51.612Z
Learning: In the lightspeed-stack repository, when provider_id == "azure", the Azure provider with provider_type "remote::azure" is guaranteed to be present in the providers list. Therefore, avoid defensive StopIteration handling for next() when locating the Azure provider in providers within src/app/endpoints/query.py. This change applies specifically to this file (or nearby provider lookup code) and relies on the invariant that the Azure provider exists; if the invariant could be violated, keep the existing StopIteration handling.
Applied to files:
src/app/endpoints/query.py
🔇 Additional comments (52)
tests/unit/models/requests/__init__.py (1)
1-1: LGTM!Docstring correctly reflects the relocated path
models.api.requests.AGENTS.md (1)
34-38: LGTM!Project structure tree accurately mirrors the new
src/models/api/requests/andsrc/models/common/layout introduced by this PR.tests/unit/models/requests/README.md (1)
1-24: LGTM!README correctly reflects the relocated model path and includes all current test file entries.
src/models/common/feedback.py (1)
1-19: LGTM!Clean enum definition with appropriate docstrings and well-defined categories.
src/models/api/requests/prompts.py (1)
1-86: LGTM!Both request models are well-structured with appropriate field constraints, docstrings, and
extra="forbid"configuration.src/app/endpoints/a2a.py (1)
47-47: Import path update looks correct and consistent with the refactor.
No concerns with this change.tests/integration/endpoints/test_streaming_query_byok_integration.py (1)
19-19: Good import migration forQueryRequestin integration tests.
This aligns with the new model module layout.tests/unit/models/requests/test_vector_store_requests.py (1)
6-6: LGTM — request model imports are updated correctly.tests/unit/app/endpoints/test_mcp_servers.py (1)
17-17: Import update is correct for MCP server request model relocation.tests/integration/endpoints/test_stream_interrupt_integration.py (1)
10-10: Looks good —StreamingInterruptRequestimport now points to the new package.tests/unit/utils/test_transcripts.py (1)
8-8: Approved —QueryRequestimport matches the newmodels.api.requestslocation.tests/unit/app/endpoints/test_feedback.py (1)
21-21: Import migration for feedback request models is correct.tests/unit/app/endpoints/test_conversations.py (1)
22-22: Looks good —ConversationUpdateRequestimport is updated correctly.tests/unit/models/requests/test_feedback_status_update_request.py (1)
3-3: Import path update is correct and consistent with the refactor.The test now targets the reorganized request-model package without changing behavior.
tests/unit/app/endpoints/test_responses.py (1)
29-29:ResponsesRequestimport migration looks good.This keeps unit tests aligned with the new
models.api.requestspackage layout.src/utils/vector_search.py (1)
21-21: Shared-model import update is correct.Using
models.common.query.SolrVectorSearchRequestfits the new shared-model split and keeps this utility aligned with the refactor.tests/integration/endpoints/test_conversations_v1_integration.py (1)
23-23: Import path migration is clean and consistent.The integration tests now reference
ConversationUpdateRequestfrom the new API request package as expected.tests/unit/models/requests/test_feedback_request.py (1)
6-7: Feedback model imports are correctly split across API/common modules.This matches the new model organization while preserving existing test intent.
tests/integration/endpoints/test_model_list.py (1)
15-15:ModelFilterimport update looks good.The test now points at the new request-model namespace with no behavioral regression risk in this change.
src/app/endpoints/stream_interrupt.py (1)
10-10: Request import update is correct.Switching to
models.api.requests.StreamingInterruptRequestis consistent with the model split and keeps endpoint typing aligned.tests/unit/app/endpoints/test_conversations_v2.py (1)
23-23:ConversationUpdateRequestimport change is good.This keeps the unit tests synchronized with the new API request-model package structure.
tests/unit/app/endpoints/test_vector_stores.py (1)
26-30: Import path update looks correct.The three vector store request model imports are consistently redirected to
models.api.requests, matching the source change insrc/app/endpoints/vector_stores.py.src/app/endpoints/feedback.py (1)
16-16: Import path update looks correct.
FeedbackRequestandFeedbackStatusUpdateRequestare now sourced frommodels.api.requests, matching the new module structure.tests/unit/app/endpoints/test_prompts.py (1)
20-20: Import path update looks correct.
PromptCreateRequestandPromptUpdateRequestare now sourced frommodels.api.requests, consistent with the endpoint source changes.tests/unit/utils/test_prompts.py (1)
9-9: Import path update looks correct.
QueryRequestis now sourced frommodels.api.requests, aligning with the consolidated API request model structure.tests/integration/endpoints/test_query_byok_integration.py (1)
18-18: Import path update looks correct.
QueryRequestis now sourced frommodels.api.requests, consistent with the rest of the BYOK test suite.src/app/endpoints/conversations_v1.py (1)
19-19: Import path update looks correct.
ConversationUpdateRequestis now sourced frommodels.api.requests, matching the new conversations request submodule.tests/integration/endpoints/test_query_integration.py (1)
21-23: Split import pattern is correct and architecturally sound.
QueryRequestfrommodels.api.requestsandAttachmentfrommodels.common.queryproperly reflect the separation of API-layer request models from shared domain types. TheAttachmentclass is correctly defined insrc/models/common/query.pyas a PydanticBaseModeland is properly exported for use in tests.src/app/endpoints/vector_stores.py (1)
25-29: Import path updates are correct.All three request models (
VectorStoreCreateRequest,VectorStoreFileCreateRequest,VectorStoreUpdateRequest) are properly defined insrc/models/api/requests/vector_stores.pyand correctly re-exported throughsrc/models/api/requests/__init__.py. No stalemodels.requestsimports remain in the codebase.tests/unit/app/endpoints/test_stream_interrupt.py (1)
10-10: LGTM – import path updated correctly.
StreamingInterruptRequestnow resolves from the newmodels.api.requestspackage, consistent with the rest of the migration.tests/unit/utils/test_vector_search.py (1)
9-9: LGTM –SolrVectorSearchRequestcorrectly moved tomodels.common.query.tests/integration/endpoints/test_health_integration.py (1)
17-17: LGTM –HealthStatuscorrectly sourced from the sharedmodels.commonpackage.tests/unit/utils/test_query.py (1)
18-25: LGTM – bothQueryRequestandAttachmentimport paths updated correctly.tests/unit/app/endpoints/test_streaming_query.py (1)
67-70: LGTM – import paths aligned with the new module structure.tests/unit/app/endpoints/test_health.py (1)
17-17: LGTM –HealthStatusandProviderHealthStatusboth correctly sourced frommodels.common.tests/integration/endpoints/test_streaming_query_integration.py (1)
15-16: LGTM – import paths updated consistently with the unit test counterparts.src/app/endpoints/query.py (1)
27-27: ⚡ Quick winNew module structure is properly wired and migration is complete.
No stale
models.requestsimports remain in the codebase. Themodels/api/requests/__init__.pypackage correctly re-exportsQueryRequest, andmodels/common/__init__.pyproperly re-exportsHealthStatus,ProviderHealthStatus,Attachment, andSolrVectorSearchRequest. All symbols are in their expected locations.src/utils/shields.py (1)
21-21: LGTM — import path updated correctly.
QueryRequestnow resolves from the newmodels.api.requestspackage, consistent with the PR-wide migration.src/models/context.py (1)
7-7: LGTM — import path updated correctly.tests/unit/app/endpoints/test_query.py (1)
15-19: LGTM — both import paths updated correctly.
QueryRequestfrommodels.api.requestsandAttachmentfrommodels.common.queryalign with the PR-wide restructuring.src/app/endpoints/streaming_query.py (1)
63-63: LGTM — import path updated correctly.src/utils/transcripts.py (1)
19-19: LGTM — import path updated correctly.src/models/api/requests/conversations.py (1)
1-22: LGTM — well-constrained model with appropriate validation.
ConversationUpdateRequestfollows all the established patterns:BaseModel, Google-style Attributes docstring, field-level constraints, andextra="forbid"to reject unknown keys.src/models/common/__init__.py (1)
1-51: LGTM — exports are consistent with the new common submodule structure.All three new re-exports (
FeedbackCategory,HealthStatus/ProviderHealthStatus,Attachment/SolrVectorSearchRequest) have matching import statements and__all__entries.src/app/endpoints/conversations_v2.py (1)
11-11: LGTM — import path correctly updated to the new module location.src/app/endpoints/mcp_servers.py (1)
14-14: Import migration verified — no stale references remain.The change at line 14 correctly uses the new absolute import path
from models.api.requests import MCPServerRegistrationRequest. A search of the codebase confirms no remaining references to the oldmodels.requestspath exist, indicating the migration is complete.docs/openapi.json (1)
11465-11465: LGTM! Documentation improvements are comprehensive and well-structured.The OpenAPI schema descriptions have been properly regenerated with refined documentation for all affected request models. Each description follows a consistent format, clearly documents attributes with appropriate detail about optionality and defaults, and aligns with the PR's objective of improving model documentation after the request model reorganization.
The backtick formatting improvement in
SolrVectorSearchRequestand the legacy compatibility note demonstrate attention to detail.Also applies to: 12590-12590, 12967-12967, 13037-13037, 14000-14000, 16453-16453, 16687-16687, 17100-17100, 19294-19294
src/app/endpoints/prompts.py (1)
16-16: LGTM — import re-pointed tomodels.api.requests.src/utils/query.py (1)
28-39: LGTM — both import paths correctly re-pointed tomodels.api.requestsandmodels.common.query.src/app/endpoints/health.py (1)
29-29: LGTM —HealthStatusandProviderHealthStatuscorrectly consolidated intomodels.common.src/app/endpoints/responses.py (1)
41-41: LGTM —ResponsesRequestimport correctly re-pointed tomodels.api.requests.src/models/api/requests/__init__.py (1)
1-30: LGTM — clean aggregator module; all 13 exports are present in both imports and__all__.
2 participants
Description
Splits requests into multiple separate files under
src/models/api/requestsor replaces shared internal models undersrc/models/common. Refines model documentations, updatesAGENTS.mdto reflect new project structure.Type of change
Tools used to create PR
Identify any AI code assistants used in this PR (for transparency and review context)
Related Tickets & Documents
Checklist before requesting a review
Testing
Summary by CodeRabbit
Refactor
Documentation