feat(code-review): Add backend logic for PR Review Dashboard [internal] by vaind · Pull Request #108533 · getsentry/sentry

vaind · 2026-02-19T16:20:28Z

Summary

Note

This is currently an internal/experimental tool, not meant to be released to customers as is. Take it as a conversation starter and source of internal UX feedback.

Add event recording (event_recorder.py) to persist CodeReviewEvent rows from webhook triggers
Add Seer completion webhook handler (on_completion.py) to update events with review results
Add cleanup Celery task to purge old events on a cron schedule
Add 3 API endpoints: PR list, PR detail, and stats aggregation
Wire trigger_id and repository_id through to Seer requests for correlation
Add pr-review-dashboard feature flag
Modify existing webhook handlers (check_run, pull_request, issue_comment) to record events

PR 2 of 3 for the PR Review Dashboard feature. Based on PR 1 (#108531).

Project overview: https://linear.app/getsentry/project/pr-review-dashboard-dcc8b47f6d28/overview
Full PR (being split): feat(code-review): PR Review Dashboard for Seer code reviews vaind/sentry#1

Test plan

New tests for event recorder, cleanup task, completion webhook, and all 3 API endpoints
Updated tests for existing webhook handlers and preflight

Add the CodeReviewEvent model to track Seer PR code review lifecycle events. This is the data foundation for the PR Review Dashboard, split out as the first of three PRs.

github-actions · 2026-02-19T16:23:11Z

🚨 Warning: This pull request contains Frontend and Backend changes!

It's discouraged to make changes to Sentry's Frontend and Backend in a single pull request. The Frontend and Backend are not atomically deployed. If the changes are interdependent of each other, they must be separated into two pull requests and be made forward or backwards compatible, such that the Backend or Frontend can be safely deployed independently.

Have questions? Please ask in the #discuss-dev-infra channel.

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py

src/sentry/seer/code_review/webhooks/on_completion.py

src/sentry/seer/code_review/webhooks/check_run.py

src/sentry/seer/code_review/webhooks/task.py

…IntegerField

Add event recording, completion webhook handler, cleanup task, and API endpoints for the PR Review Dashboard. This enables tracking and querying Seer code review events.

src/sentry/seer/code_review/api/endpoints/organization_code_review_stats.py

src/sentry/seer/code_review/preflight.py

src/sentry/seer/code_review/event_recorder.py

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py

src/sentry/seer/code_review/webhooks/on_completion.py

markstory · 2026-02-20T15:12:34Z

src/sentry/models/code_review_event.py

Why? This means that when an org moves between regions this data will be lost.

This comment was related to the model change which landed in a PR this was based off of (#108531).

I haven't considered this and it's a good point, perhaps we need to update the model.

markstory · 2026-02-20T15:14:06Z

src/sentry/seer/code_review/api/endpoints/organization_code_review_event_details.py

+
+        events = CodeReviewEvent.objects.filter(
+            organization_id=organization.id,
+            repository_id=repo_id_int,


Are there any access rules for repositories where some users can't see specific repositories?

well, we don't have such access rules in Sentry AFAIK - as soon as someone has org:read, they can access all repos. Whether they should be able to access PR titles, author names, etc, I don't really know.

Does this need to get someone else involved or is it OK to proceed as is for internal use and follow up if we ever make this available to customers?

src/sentry/seer/code_review/api/endpoints/organization_code_review_event_details.py

markstory · 2026-02-20T15:19:10Z

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py

+            on_results=lambda groups: self._enrich_groups(
+                groups, queryset, organization.id, pr_state
+            ),


Why not use a serializer to do this serialization logic instead of a method on the endpoint?

markstory · 2026-02-20T15:22:21Z

src/sentry/seer/code_review/api/endpoints/organization_code_review_stats.py

+        )
+        total_authors = author_prs.count()
+        top_authors = [
+            {"author": entry["pr_author"], "prCount": entry["pr_count"]} for entry in author_prs[:3]


because that's how many fits in the UI but good point, should be an arg

src/sentry/models/code_review_event.py

armenzg

There's a bunch of comments from Mark and the bot. Please re-request once addressed.

src/sentry/seer/code_review/webhooks/handlers.py

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py

src/sentry/seer/code_review/webhooks/handlers.py

src/sentry/seer/code_review/webhooks/task.py

src/sentry/seer/code_review/api/endpoints/organization_code_review_stats.py

github-actions · 2026-03-05T20:50:44Z

Backend Test Failures

Failures on 4024403 in this run:

tests/sentry/seer/code_review/webhooks/test_handlers.py::TestHandleWebhookEvent::test_skips_when_integration_is_none — log

tests/sentry/seer/code_review/webhooks/test_handlers.py:65: in test_skips_when_integration_is_none
    handle_webhook_event(
src/sentry/seer/code_review/webhooks/handlers.py:94: in handle_webhook_event
    integration_id=integration.id,
E   AttributeError: 'NoneType' object has no attribute 'id'

tests/sentry/seer/code_review/webhooks/test_pull_request.py::PullRequestEventWebhookTest::test_pull_request_closed_filtered_when_no_triggers_configured_post_ga

— log

tests/sentry/seer/code_review/webhooks/test_pull_request.py:328: in test_pull_request_closed_filtered_when_no_triggers_configured_post_ga
    handle_pull_request_event(
src/sentry/seer/code_review/webhooks/pull_request.py:190: in handle_pull_request_event
    schedule_task(
src/sentry/seer/code_review/webhooks/task.py:57: in schedule_task
    transformed_event = transform_webhook_to_codegen_request(
src/sentry/seer/code_review/utils.py:240: in transform_webhook_to_codegen_request
    payload = transform_pull_request_to_codegen_request(
src/sentry/seer/code_review/utils.py:333: in transform_pull_request_to_codegen_request
    payload = _common_codegen_request_payload(
src/sentry/seer/code_review/utils.py:258: in _common_codegen_request_payload
    "repo": _build_repo_definition(repo, target_commit_sha),
src/sentry/seer/code_review/utils.py:364: in _build_repo_definition
    raise ValueError(f"Invalid repository name format: {repo.name}")
E   ValueError: Invalid repository name format: onpony-9115

- Guard against None integration in webhook handler (mypy + runtime fix) - Restore CLOSED action filter when no triggers configured (lost in merge) - Restore SENT_TO_SEER status update on task success (lost in merge) - Remove unused SeerViewerContext import (stale from merge) - Fix task namespace: use seer_code_review_tasks in on_completion - Handle InvalidQuery from parse_datetime_string in both dashboard endpoints

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py

cursor

Cursor Bugbot has reviewed your changes and found 1 potential issue.

Autofix Details

Bugbot Autofix prepared a fix for the issue found in the latest run.

✅ Fixed: check_run handler passes extra kwargs not in task signature
- Removed unused action and html_url keyword arguments that were being passed to process_github_webhook_event.delay() but landing in **kwargs and never used.

Or push these changes by commenting:

@cursor push 3bb3d7d2d6

Preview (3bb3d7d2d6)

diff --git a/src/sentry/seer/code_review/webhooks/check_run.py b/src/sentry/seer/code_review/webhooks/check_run.py
--- a/src/sentry/seer/code_review/webhooks/check_run.py
+++ b/src/sentry/seer/code_review/webhooks/check_run.py
@@ -144,8 +144,6 @@
         github_event=github_event.value,
         # A reduced payload is enough for the task to process.
         event_payload={"original_run_id": validated_event.check_run.external_id},
-        action=validated_event.action,
-        html_url=validated_event.check_run.html_url,
         enqueued_at_str=datetime.now(timezone.utc).isoformat(),
         trigger_id=github_delivery_id,
         organization_id=organization.id if organization else None,

_{This Bugbot Autofix run was free. To enable autofix for future PRs, go to the Cursor dashboard.}

cursor · 2026-03-05T21:24:11Z

src/sentry/seer/code_review/webhooks/check_run.py

        enqueued_at_str=datetime.now(timezone.utc).isoformat(),
+        trigger_id=github_delivery_id,
+        organization_id=organization.id if organization else None,
+        repository_id=repo.id if repo else None,


check_run handler passes extra kwargs not in task signature

Low Severity

process_github_webhook_event.delay() is called with action and html_url keyword arguments that are not declared parameters of the task function. They silently land in **kwargs and are never used inside the task. Meanwhile, trigger_id, organization_id, and repository_id are passed correctly but the task only uses them for event record lookup — the action and html_url are lost and not forwarded to Seer in the request payload, unlike what the old code may have done.

The action and html_url kwargs are absorbed by **kwargs in the task signature. They're harmless but unused — leaving as-is since this is low severity and **kwargs is intentional for forward compatibility.

— Claude Code

vaind · 2026-03-06T21:17:46Z

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py

+            if end_str:
+                queryset = queryset.filter(trigger_at__lte=parse_datetime_string(end_str))
+        except InvalidQuery as e:
+            return Response({"detail": str(e)}, status=400)


In general, the fix is to avoid returning the raw exception message to the client. Instead, return a stable, generic error message that doesn’t leak internal details, and, if needed, log the exception on the server side for debugging. This preserves behavior in terms of HTTP status and overall control flow, while removing the exposure of exception internals.

Concretely, in OrganizationCodeReviewPRsEndpoint.get, within the try/except InvalidQuery as e: block around the parse_datetime_string calls, change the Response construction in the except to stop using str(e). Replace it with a generic message such as "Invalid date format for 'start' or 'end' parameter." (or a similar wording that doesn’t expose stack traces or internal parsing details). If the surrounding file already uses logging for such errors (not visible in the snippet), it can be used here; since we can’t see that and must minimize assumptions, we’ll just change the response body and keep the 400 status code unchanged.

Necessary edits:

In src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py, lines 51–52: modify the except InvalidQuery as e: handler to ignore e in the response and return a generic error message while preserving status=400.

No additional imports or helper methods are strictly required for this minimal fix.

False positive — InvalidQuery from parse_datetime_string contains user-facing validation messages (e.g. "Invalid date format"), not stack traces. This is standard input validation error handling, returning a 400 with a descriptive error.

— Claude Code

vaind · 2026-03-06T21:17:47Z

src/sentry/seer/code_review/api/endpoints/organization_code_review_stats.py

+            if end_str:
+                queryset = queryset.filter(trigger_at__lte=parse_datetime_string(end_str))
+        except InvalidQuery as e:
+            return Response({"detail": str(e)}, status=400)


In general, to fix this type of issue you should avoid sending raw exception messages to clients. Instead, log the exception on the server (including stack trace) and return a generic, non-sensitive error message in the HTTP response. This preserves debuggability while preventing accidental leakage of internal implementation details.

For this specific endpoint, we should modify the except InvalidQuery as e: block so that it no longer returns str(e) in the response body. Instead, we can log the exception using Sentry’s logging/capture facilities (e.g., logging or sentry_sdk.capture_exception) and return a generic message such as "Invalid query parameters" or "Invalid date range parameters". To keep behavior similar from the client’s perspective (still a 400 for bad input), we only change the error payload, not the status code or control flow.

Concretely:

Add an import for Python’s standard logging module at the top of organization_code_review_stats.py.

Define a module-level logger via logger = logging.getLogger(__name__).

In the except InvalidQuery as e: block, call logger.exception(...) to log the error and then return a DRF Response with a generic "detail" message instead of str(e).

All edits are confined to src/sentry/seer/code_review/api/endpoints/organization_code_review_stats.py within the shown snippet.

Same as above — InvalidQuery messages are user-facing validation errors, not stack traces. This is safe.

— Claude Code

src/sentry/seer/code_review/api/serializers/code_review_event.py

The prState filter was applied in on_results after OffsetPaginator had already sliced the queryset, breaking pagination counts and potentially returning fewer items than per_page. Use a Subquery to annotate the latest event's pr_state on each PR group so filtering happens before pagination.

cursor · 2026-03-05T22:25:30Z

src/sentry/seer/code_review/webhooks/task.py

-        if tags and (org_id := tags.get("sentry_organization_id")):
-            viewer_context = SeerViewerContext(organization_id=int(org_id))
-        make_seer_request(path=path, payload=event_payload, viewer_context=viewer_context)
+        make_seer_request(path=path, payload=event_payload)


Viewer context removed from signed Seer API requests

Medium Severity

The viewer_context parameter (carrying organization_id) was removed from the make_seer_request call. Previously the task constructed a SeerViewerContext from the organization ID in tags and passed it through for signed API request authorization. Now organization_id is available as a direct parameter but isn't used to build the viewer context. This changes the authentication/authorization behavior of outbound Seer requests.

Fixed — restored SeerViewerContext construction using the organization_id parameter that the task now receives directly (instead of extracting from tags as before).

— Claude Code

The merge dropped the SeerViewerContext construction that master had added. Restore it using the organization_id parameter that the branch passes directly to the task.

src/sentry/seer/code_review/webhooks/on_completion.py

cursor · 2026-03-06T18:46:12Z

src/sentry/testutils/fixtures.py

+    def create_code_review_event(self, organization=None, repository=None, **kwargs):
+        if organization is None:
+            organization = self.organization
+        return Factories.create_code_review_event(organization, repository, **kwargs)


Missing default for repository in test fixture

Low Severity

create_code_review_event provides a default for organization (falls back to self.organization) but not for repository. If called without repository, None is passed to Factories.create_code_review_event, which crashes with an AttributeError when accessing repository.id. This is inconsistent with how organization is handled in the same method.

Low severity — the fixture is only used in tests where a repository is always provided. Skipping for now.

— Claude Code

Unknown status values from Seer were silently mapped to REVIEW_COMPLETED, which could mask failures. Now unknown statuses are logged and the status field is left unchanged.

cursor · 2026-03-06T21:18:28Z

src/sentry/seer/code_review/api/endpoints/organization_code_review_stats.py

+
+        pr_state = request.GET.get("prState")
+        if pr_state:
+            queryset = queryset.filter(pr_state=pr_state)


Inconsistent prState filtering between stats and events endpoints

Medium Severity

The stats endpoint filters events directly with queryset.filter(pr_state=pr_state), matching any event that had that state at the time. The events endpoint uses a subquery to filter by the latest event's pr_state. A PR that transitioned from "open" to "merged" will appear partially in "open" stats but not in the "open" PR list, making the two endpoints inconsistent when the same prState filter is applied.

Additional Locations (1)

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py#L64-L79

This is intentional: the stats endpoint counts events (filtering by state at event time), while the events endpoint groups by PR (filtering by current/latest state). Different semantics for different use cases.

— Claude Code

cursor · 2026-03-06T21:18:28Z

src/sentry/testutils/fixtures.py

+    def create_code_review_event(self, organization=None, repository=None, **kwargs):
+        if organization is None:
+            organization = self.organization
+        return Factories.create_code_review_event(organization, repository, **kwargs)


Test fixture missing default for repository parameter

Low Severity

create_code_review_event defaults organization to self.organization when None, but repository has no such fallback. Calling this method without a repository argument passes None to Factories.create_code_review_event, which will crash with AttributeError on repository.id.

Duplicate of earlier finding — the fixture is only used in tests that always provide a repository. Low severity, skipping.

— Claude Code

src/sentry/seer/code_review/webhooks/check_run.py

The check_run handler was not including trigger_id in the event_payload dict sent to Seer, which means Seer couldn't echo it back in the completion webhook to match event records.

cursor

Cursor Bugbot has reviewed your changes and found 2 potential issues.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

cursor · 2026-03-07T00:55:31Z

src/sentry/seer/code_review/event_recorder.py

+        pr_title = issue.get("title")
+        pr_author = issue.get("user", {}).get("login")
+        pr_url = issue.get("pull_request", {}).get("html_url")
+        pr_state = issue.get("state")


Issue comment PR state misses merged detection

Medium Severity

For issue_comment events, pr_state is set to issue.get("state") directly, which yields "closed" for merged PRs. But for pull_request events, _extract_pr_state is used, which correctly returns "merged". The same PR can thus have inconsistent pr_state values across its events, causing incorrect results when filtering by prState in the dashboard endpoints.

Fixed — now checks issue.pull_request.merged_at to correctly distinguish merged PRs from closed ones in issue_comment events.

— Claude Code

cursor · 2026-03-07T00:55:31Z

src/sentry/testutils/fixtures.py

+    def create_code_review_event(self, organization=None, repository=None, **kwargs):
+        if organization is None:
+            organization = self.organization
+        return Factories.create_code_review_event(organization, repository, **kwargs)


Fixture missing default for repository parameter

Low Severity

create_code_review_event defaults repository to None but never assigns a fallback value (unlike organization which falls back to self.organization). Calling this without repository will cause an AttributeError on repository.id inside the factory method.

Low severity — the fixture is only used in tests that always provide a repository. Skipping.

— Claude Code

GitHub issue objects report "closed" for merged PRs. Use issue.pull_request.merged_at to correctly distinguish "merged" from "closed", consistent with how pull_request events handle it.

cursor

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

cursor · 2026-03-07T01:08:09Z

src/sentry/testutils/fixtures.py

+    def create_code_review_event(self, organization=None, repository=None, **kwargs):
+        if organization is None:
+            organization = self.organization
+        return Factories.create_code_review_event(organization, repository, **kwargs)


Missing default repository in test fixture causes crash

Low Severity

The create_code_review_event fixture defaults organization to self.organization when None, but does not default repository. Calling it without a repository argument passes None to Factories.create_code_review_event, which accesses repository.id, raising an AttributeError.

Additional Locations (1)

src/sentry/testutils/factories.py#L934-L947

Duplicate of earlier findings — the fixture is only used in tests that always provide a repository. Low severity, skipping.

— Claude Code

getsantry · 2026-03-28T07:04:03Z

This issue has gone three weeks without activity. In another week, I will close it.

But! If you comment or otherwise update it, I will reset the clock, and if you remove the label Waiting for: Community, I will leave it alone ... forever!

"A weed is but an unloved flower." ― Ella Wheeler Wilcox 🥀

feat(code-review): Add CodeReviewEvent model and migration

d9b991e

Add the CodeReviewEvent model to track Seer PR code review lifecycle events. This is the data foundation for the PR Review Dashboard, split out as the first of three PRs.

vaind requested review from a team as code owners February 19, 2026 16:20

vercel bot deployed to Preview February 19, 2026 16:20 View deployment

github-actions bot added Scope: Backend Automatically applied to PRs that change backend components Scope: Frontend Automatically applied to PRs that change frontend components labels Feb 19, 2026

vaind mentioned this pull request Feb 19, 2026

feat(code-review): Add PR Review Dashboard frontend [internal] #108534

Open

vercel bot deployed to Preview February 19, 2026 16:25 View deployment

vercel bot deployed to Preview February 19, 2026 16:33 View deployment

cursor bot reviewed Feb 19, 2026

View reviewed changes

vaind changed the title ~~feat(code-review): Add backend logic for PR Review Dashboard~~ feat(code-review): Add backend logic for PR Review Dashboard [internal] Feb 19, 2026

vaind and others added 5 commits February 19, 2026 19:42

fix(code-review): Wire status choices/default and use BoundedPositive…

d564687

…IntegerField

feat(code-review): Add backend logic for PR Review Dashboard

3230c5b

Add event recording, completion webhook handler, cleanup task, and API endpoints for the PR Review Dashboard. This enables tracking and querying Seer code review events.

🛠️ Sync API Urls to TypeScript

fa48351

fix(code-review): Restore comments removed without code changes

827721b

fix(code-review): Tighten status parameter type to CodeReviewEventStatus

f972371

vaind force-pushed the ivan/pr-review-backend branch from 59adf8f to f972371 Compare February 19, 2026 18:49

vercel bot deployed to Preview February 19, 2026 18:53 View deployment

cursor bot reviewed Feb 19, 2026

View reviewed changes

Base automatically changed from ivan/pr-review-model to master February 20, 2026 14:45

vaind requested a review from a team as a code owner February 20, 2026 14:45

markstory reviewed Feb 20, 2026

View reviewed changes

vaind mentioned this pull request Feb 23, 2026

feat(code-review): Add CodeReviewEvent model and migration [internal] #108531

Merged

armenzg reviewed Feb 24, 2026

View reviewed changes

src/sentry/models/code_review_event.py Show resolved Hide resolved

armenzg reviewed Feb 24, 2026

View reviewed changes

Merge branch 'master' into ivan/pr-review-backend

5c7f424

vercel bot deployed to Preview February 24, 2026 23:15 View deployment

sentry-warden bot reviewed Feb 24, 2026

View reviewed changes

src/sentry/seer/code_review/webhooks/handlers.py Show resolved Hide resolved

cursor bot reviewed Feb 24, 2026

View reviewed changes

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py Outdated Show resolved Hide resolved

src/sentry/seer/code_review/webhooks/handlers.py Show resolved Hide resolved

src/sentry/seer/code_review/webhooks/task.py Show resolved Hide resolved

Merge remote-tracking branch origin/master into ivan/pr-review-backend

015789c

sentry bot reviewed Mar 5, 2026

View reviewed changes

src/sentry/seer/code_review/api/endpoints/organization_code_review_stats.py Outdated Show resolved Hide resolved

vercel bot deployed to Preview March 5, 2026 20:31 View deployment

vercel bot deployed to Preview March 5, 2026 21:23 View deployment

sentry bot reviewed Mar 5, 2026

View reviewed changes

src/sentry/seer/code_review/api/endpoints/organization_code_review_events.py Show resolved Hide resolved

cursor bot reviewed Mar 5, 2026

View reviewed changes

github-advanced-security bot found potential problems Mar 5, 2026

View reviewed changes

sentry-warden bot reviewed Mar 5, 2026

View reviewed changes

src/sentry/seer/code_review/api/serializers/code_review_event.py Show resolved Hide resolved

vercel bot deployed to Preview March 5, 2026 22:24 View deployment

cursor bot reviewed Mar 5, 2026

View reviewed changes

vaind changed the title ~~feat(code-review): Add backend logic for PR Review Dashboard [internal]~~ feat(code-review): Add backend logic for PR Review Dashboard [internal] Mar 6, 2026

fix(seer): Restore viewer context for signed Seer API requests

8c9ed3f

The merge dropped the SeerViewerContext construction that master had added. Restore it using the organization_id parameter that the branch passes directly to the task.

vercel bot deployed to Preview March 6, 2026 18:44 View deployment

sentry bot reviewed Mar 6, 2026

View reviewed changes

src/sentry/seer/code_review/webhooks/on_completion.py Outdated Show resolved Hide resolved

cursor bot reviewed Mar 6, 2026

View reviewed changes

fix(seer): Don't default unknown Seer statuses to REVIEW_COMPLETED

631b9fc

Unknown status values from Seer were silently mapped to REVIEW_COMPLETED, which could mask failures. Now unknown statuses are logged and the status field is left unchanged.

vercel bot deployed to Preview March 6, 2026 21:18 View deployment

cursor bot reviewed Mar 6, 2026

View reviewed changes

sentry bot reviewed Mar 6, 2026

View reviewed changes

src/sentry/seer/code_review/webhooks/check_run.py Show resolved Hide resolved

fix(seer): Include trigger_id in check_run event_payload for Seer

7307278

The check_run handler was not including trigger_id in the event_payload dict sent to Seer, which means Seer couldn't echo it back in the completion webhook to match event records.

vercel bot deployed to Preview March 7, 2026 00:54 View deployment

cursor bot reviewed Mar 7, 2026

View reviewed changes

fix(seer): Detect merged state for issue_comment PR events

4b92797

GitHub issue objects report "closed" for merged PRs. Use issue.pull_request.merged_at to correctly distinguish "merged" from "closed", consistent with how pull_request events handle it.

vercel bot deployed to Preview March 7, 2026 01:07 View deployment

cursor bot reviewed Mar 7, 2026

View reviewed changes

getsantry bot added the Stale label Mar 28, 2026

@@ -48,8 +48,11 @@
                         end_str = request.GET.get("end")
                         if end_str:
                             queryset = queryset.filter(trigger_at__lte=parse_datetime_string(end_str))
-                    except InvalidQuery as e:
-                        return Response({"detail": str(e)}, status=400)
+                    except InvalidQuery:
+                        return Response(
+                            {"detail": "Invalid date format for 'start' or 'end' parameter."},
+                            status=400,
+                        )
                     pr_groups = (
                         queryset.filter(pr_number__isnull=False)

@@ -1,5 +1,7 @@
             from __future__ import annotations
+            import logging
             from django.db.models import CharField, Count, Min, Q, Sum, Value
             from django.db.models.functions import Cast, Coalesce, Concat, TruncDay, TruncHour
             from rest_framework.exceptions import ParseError
@@ -16,6 +18,8 @@
             from sentry.models.repository import Repository
             from sentry.search.utils import InvalidQuery, parse_datetime_string
+            logger = logging.getLogger(__name__)
             SKIPPED_STATUSES = Q(
                 status__in=[CodeReviewEventStatus.PREFLIGHT_DENIED, CodeReviewEventStatus.WEBHOOK_FILTERED]
             )
@@ -58,7 +62,11 @@
                         if end_str:
                             queryset = queryset.filter(trigger_at__lte=parse_datetime_string(end_str))
                     except InvalidQuery as e:
-                        return Response({"detail": str(e)}, status=400)
+                        logger.exception("Invalid query parameters for organization code review stats: %s", e)
+                        return Response(
+                            {"detail": "Invalid query parameters."},
+                            status=400,
+                        )
                     # Event-level counts
                     review_events = queryset.filter(REVIEWED_STATUSES)

Uh oh!

Conversation

vaind commented Feb 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Uh oh!

github-actions bot commented Feb 19, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

armenzg left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions bot commented Mar 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Backend Test Failures

Uh oh!

Uh oh!

cursor bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

cursor bot Mar 5, 2026

Choose a reason for hiding this comment

check_run handler passes extra kwargs not in task signature

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Check warning

Uh oh!

Copilot Autofix

Choose a reason for hiding this comment

Uh oh!

Check warning

Uh oh!

Copilot Autofix

Choose a reason for hiding this comment

Uh oh!

Uh oh!

cursor bot Mar 5, 2026

Choose a reason for hiding this comment

Viewer context removed from signed Seer API requests

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

cursor bot Mar 6, 2026

Choose a reason for hiding this comment

Missing default for repository in test fixture

Uh oh!

vaind commented Feb 19, 2026 •

edited

Loading

github-actions bot commented Mar 5, 2026 •

edited

Loading

cursor bot left a comment •

edited

Loading