DeepSeek V4.1 Flash · Heavy

replays-self-serve-bulk-delete-system-sentry

Sentry · Python · M · 100–299 lines · 3 files

The generalist plus 9 specialist lens agents in parallel, then the same verify and dedup steps. Judge: Claude Haiku 5.5.

Gold bugs found
0 of 10
Comments published
7
Comments matching a bug
0
Comments not matching
7
Cost of this PR
$0.602

Published comments (7)

highbugstatic/app/views/dashboards/widgetCard/chart.tsx:164-174

Table widget flag branch renders blank grid with empty data — Every table widget renders a blank grid once the flag is enabled, since `fields` and `result` are discarded. Feed the already-computed `result.data`, `result.meta` and `fields` into `TableWidgetVisualization`.

mediumbugsrc/sentry/replays/endpoints/project_replay_summarize_breadcrumbs.py:79-84

Error context dropped when field params are provided — Any caller passing field params gets [] for error_ids, so the new error context is silently dropped. Request the replay fields independently of the client's field list so error ids are always available.

mediumbugsrc/sentry/replays/endpoints/project_replay_summarize_breadcrumbs.py:79-82

Invalid field query param causes 500 in replay breadcrumbs — An authenticated `?field=bogus` request raises `KeyError` in `process_raw_response` and returns an unhandled 500 instead of a 400. Filter the requested fields against the known response keys before calling it.

mediumbugtests/sentry/api/endpoints/test_browser_reporting_collector.py:64-70

Report tests no longer post with application/reports+json content type — Without the browser-reporting content type DRF uses the generic `JSONParser`, so these tests would still pass if `BrowserReportsJSONParser` broke. Restore `content_type="application/reports+json"` on at least one request.

mediumsecuritysrc/sentry/issues/endpoints/browser_reporting_collector.py:110-114

Unauthenticated endpoint logs raw unvalidated report payload — This unauthenticated, CSRF-exempt endpoint lets anyone persist arbitrary content, and browser reports carry URLs, tokens and page messages into a retained warning log. Log only non-sensitive shape information instead of `raw_report`.

mediumbugsrc/sentry/replays/endpoints/project_replay_summarize_breadcrumbs.py:107-120

Replay error events mispaired by positional get_multi zip — `NodeStorage.get_multi` dedupes ids and appends cache hits after service results, so `zip` mispairs ids with another event's data and drops trailing ones. Look events up by node id instead of relying on positional order.

mediumbugsrc/sentry/issues/endpoints/browser_reporting_collector.py:50-60

Mutually exclusive age/timestamp check uses truthiness, so 0 slips through — A legal value of 0 is falsy, so `timestamp`/`age` pairs like `{"timestamp": 0, "age": 5}` pass the mutually exclusive check. Use existence checks such as `"age" in self.initial_data` in `validate_timestamp` and `validate_age`.

Gold bugs found (0)

None.

Gold bugs missed (10)

mediumThe age/timestamp mutual-exclusion validators check truthiness via initial_data.get(), so a zero value is treated as absent and a report carrying both fields passes validation.
mediumThe validation-failure warning logs the entire raw, attacker-controlled browser report instead of a redacted subset.
mediumThe Snuba replay lookup for error IDs runs even when enable_error_context=false, adding a needless query (and failure point) to opt-out requests.
highWith the use-table-widget-visualization flag on, the widget renders TableWidgetVisualization with hardcoded empty columns and data instead of the query results, so every table widget shows an empty table.
mediumThe custom cell renderers fall back to default rendering on any falsy return (0, '', null), not only undefined, so a renderer cannot render zero or an empty cell.
mediumThe attribute-keys query key no longer includes the organization slug, so two organizations with the same filters share a cache entry and switching orgs can show the previous org's attribute names.
mediumEvery paginated breadcrumb-summary request queries the replay and fetches all of its error events from nodestore with no cap, adding latency and memory proportional to the replay's error count on each page.
mediumHIDDEN_ATTRIBUTES is applied only when a search query is present; the no-search early return shows all attributes, so the hidden attributes still appear in the default span attributes view.
mediumThe bulk delete test sets max_segment_id to None to cover the new guard in _make_recording_filenames, but it mocks delete_matched_rows, so the guard is never executed and removing it would not fail the test.
mediumError event timestamps (seconds) are compared with replay recording event timestamps (milliseconds), so errors are not interleaved chronologically with breadcrumbs.