feat(explorer): support generic table and timeseries rpcs and migrate existing tools to them (#103413)

Allows us to make generic (dataset agnostic) EAP table and timeseries
queries through Seer RPC. These 2 RPCs are mainly wrappers around the
/events/ and /events-stats/ sentry endpoints. Bonus of supporting
project_slug filters, which is what the agent works with

For testing purposes, the trace_query RPCs are still supported but call
the generic ones under the hood. We plan to eventually deprecate these
and migrate the agent tools to the generics

Author: aliu39

src/sentry/seer/endpoints/seer_rpc.py

@@ -86,6 +86,8 @@
     rpc_get_transactions_for_project,
 )
 from sentry.seer.explorer.tools import (
+    execute_table_query,
+    execute_timeseries_query,
     execute_trace_query_chart,
     execute_trace_query_table,
     get_issue_details,
@@ -1199,6 +1201,8 @@ def check_repository_integrations_status(*, repository_integrations: list[dict[s
     "get_issue_details": get_issue_details,
     "execute_trace_query_chart": execute_trace_query_chart,
     "execute_trace_query_table": execute_trace_query_table,
+    "execute_table_query": execute_table_query,
+    "execute_timeseries_query": execute_timeseries_query,
     "get_repository_definition": get_repository_definition,
     "call_custom_tool": call_custom_tool,
     #

src/sentry/seer/explorer/tools.py

@@ -5,11 +5,12 @@
 
 from sentry import eventstore, features
 from sentry.api import client
+from sentry.api.endpoints.organization_events_timeseries import TOP_EVENTS_DATASETS
 from sentry.api.serializers.base import serialize
 from sentry.api.serializers.models.event import EventSerializer, IssueEventSerializerResponse
 from sentry.api.serializers.models.group import GroupSerializer
 from sentry.api.utils import default_start_end_dates
-from sentry.constants import ObjectStatus
+from sentry.constants import ALL_ACCESS_PROJECT_ID, ObjectStatus
 from sentry.models.apikey import ApiKey
 from sentry.models.group import Group
 from sentry.models.organization import Organization
@@ -18,14 +19,15 @@
 from sentry.replays.post_process import process_raw_response
 from sentry.replays.query import query_replay_id_by_prefix, query_replay_instance
 from sentry.search.eap.types import SearchResolverConfig
-from sentry.search.events.types import SnubaParams
+from sentry.search.events.types import SAMPLING_MODES, SnubaParams
 from sentry.seer.autofix.autofix import get_all_tags_overview
 from sentry.seer.constants import SEER_SUPPORTED_SCM_PROVIDERS
 from sentry.seer.sentry_data_models import EAPTrace
 from sentry.services.eventstore.models import Event, GroupEvent
 from sentry.snuba.referrer import Referrer
 from sentry.snuba.spans_rpc import Spans
 from sentry.snuba.trace import query_trace_data
+from sentry.snuba.utils import get_dataset
 from sentry.utils.dates import parse_stats_period
 
 logger = logging.getLogger(__name__)
@@ -175,6 +177,151 @@ def execute_trace_query_table(
     return resp.data
 
 
+def execute_table_query(
+    *,
+    org_id: int,
+    dataset: str,
+    fields: list[str],
+    query: str,
+    sort: str,
+    per_page: int,
+    stats_period: str,
+    project_ids: list[int] | None = None,
+    project_slugs: list[str] | None = None,
+    sampling_mode: SAMPLING_MODES = "NORMAL",
+) -> dict[str, Any] | None:
+    """
+    Execute a query to get table data by calling the events endpoint.
+
+    Arg notes:
+        project_ids: The IDs of the projects to query. Cannot be provided with project_slugs.
+        project_slugs: The slugs of the projects to query. Cannot be provided with project_ids.
+        If neither project_ids nor project_slugs are provided, all active projects will be queried.
+    """
+    try:
+        organization = Organization.objects.get(id=org_id)
+    except Organization.DoesNotExist:
+        logger.warning("Organization not found", extra={"org_id": org_id})
+        return None
+
+    if not project_ids and not project_slugs:
+        project_ids = [ALL_ACCESS_PROJECT_ID]
+    # Note if both project_ids and project_slugs are provided, the API request will 400.
+
+    params: dict[str, Any] = {
+        "dataset": dataset,
+        "field": fields,
+        "query": query,
+        "sort": sort if sort else ("-timestamp" if "timestamp" in fields else None),
+        "per_page": per_page,
+        "statsPeriod": stats_period,
+        "project": project_ids,
+        "projectSlug": project_slugs,
+        "sampling": sampling_mode,
+        "referrer": Referrer.SEER_RPC,
+    }
+
+    # Remove None values
+    params = {k: v for k, v in params.items() if v is not None}
+
+    # Call sentry API client. This will raise API errors for non-2xx / 3xx status.
+    resp = client.get(
+        auth=ApiKey(organization_id=organization.id, scope_list=["org:read", "project:read"]),
+        user=None,
+        path=f"/organizations/{organization.slug}/events/",
+        params=params,
+    )
+    return resp.data
+
+
+def execute_timeseries_query(
+    *,
+    org_id: int,
+    dataset: str,
+    y_axes: list[str],
+    group_by: list[str] | None = None,
+    query: str,
+    stats_period: str,
+    interval: str | None = None,
+    project_ids: list[int] | None = None,
+    project_slugs: list[str] | None = None,
+    sampling_mode: SAMPLING_MODES = "NORMAL",
+    partial: Literal["0", "1"] | None = None,
+) -> dict[str, Any] | None:
+    """
+    Execute a query to get chart/timeseries data by calling the events-stats endpoint.
+
+    Arg notes:
+        interval: The interval of each bucket. Valid stats period format, e.g. '3h'.
+        partial: Whether to allow partial buckets if the last bucket does not align with rollup.
+        project_ids: The IDs of the projects to query. Cannot be provided with project_slugs.
+        project_slugs: The slugs of the projects to query. Cannot be provided with project_ids.
+        If neither project_ids nor project_slugs are provided, all active projects will be queried.
+    """
+    try:
+        organization = Organization.objects.get(id=org_id)
+    except Organization.DoesNotExist:
+        logger.warning("Organization not found", extra={"org_id": org_id})
+        return None
+
+    group_by = group_by or []
+    if not project_ids and not project_slugs:
+        project_ids = [ALL_ACCESS_PROJECT_ID]
+    # Note if both project_ids and project_slugs are provided, the API request will 400.
+
+    params: dict[str, Any] = {
+        "dataset": dataset,
+        "yAxis": y_axes,
+        "field": y_axes + group_by,
+        "query": query,
+        "statsPeriod": stats_period,
+        "interval": interval,
+        "project": project_ids,
+        "projectSlug": project_slugs,
+        "sampling": sampling_mode,
+        "referrer": Referrer.SEER_RPC,
+        "partial": partial,
+        "excludeOther": "0",  # Always include "Other" series
+    }
+
+    # Add top_events if group_by is provided
+    if group_by and get_dataset(dataset) in TOP_EVENTS_DATASETS:
+        params["topEvents"] = 5
+
+    # Remove None values
+    params = {k: v for k, v in params.items() if v is not None}
+
+    # Call sentry API client. This will raise API errors for non-2xx / 3xx status.
+    resp = client.get(
+        auth=ApiKey(organization_id=organization.id, scope_list=["org:read", "project:read"]),
+        user=None,
+        path=f"/organizations/{organization.slug}/events-stats/",
+        params=params,
+    )
+    data = resp.data
+
+    # Always normalize to the nested {"metric": {"data": [...]}} format for consistency
+    metric_is_single = len(y_axes) == 1
+    metric_name = y_axes[0] if metric_is_single else None
+    if metric_name and metric_is_single:
+        # Handle grouped data with single metric: wrap each group's data in the metric name
+        if group_by:
+            return {
+                group_value: (
+                    {metric_name: group_data}
+                    if isinstance(group_data, dict) and "data" in group_data
+                    else group_data
+                )
+                for group_value, group_data in data.items()
+            }
+
+        # Handle non-grouped data with single metric: wrap data in the metric name
+        if isinstance(data, dict) and "data" in data:
+            return {metric_name: data}
+
+    return data
+
+
 def get_trace_waterfall(trace_id: str, organization_id: int) -> EAPTrace | None:
     """
     Get the full span waterfall and connected errors for a trace.
@@ -352,7 +499,8 @@ def _get_issue_event_timeseries(
     first_seen_delta: timedelta,
 ) -> tuple[dict[str, Any], str, str] | None:
     """
-    Get event counts over time for an issue by calling the events-stats endpoint.
+    Get event counts over time for an issue (no group by) by calling the events-stats endpoint. Dynamically picks
+    a stats period and interval based on the issue's first seen date and EVENT_TIMESERIES_RESOLUTIONS.
     """
 
     stats_period, interval = None, None
@@ -364,35 +512,21 @@ def _get_issue_event_timeseries(
     stats_period = stats_period or "90d"
     interval = interval or "3d"
 
-    params: dict[str, Any] = {
-        "dataset": "issuePlatform",
-        "query": f"issue:{issue_short_id}",
-        "yAxis": "count()",
-        "partial": "1",
-        "statsPeriod": stats_period,
-        "interval": interval,
-        "project": project_id,
-        "referrer": Referrer.SEER_RPC,
-    }
-
-    resp = client.get(
-        auth=ApiKey(organization_id=organization.id, scope_list=["org:read", "project:read"]),
-        user=None,
-        path=f"/organizations/{organization.slug}/events-stats/",
-        params=params,
+    data = execute_timeseries_query(
+        org_id=organization.id,
+        dataset="issuePlatform",
+        y_axes=["count()"],
+        group_by=[],
+        query=f"issue:{issue_short_id}",
+        stats_period=stats_period,
+        interval=interval,
+        project_ids=[project_id],
+        partial="1",
     )
-    if resp.status_code != 200 or not (resp.data or {}).get("data"):
-        logger.warning(
-            "Failed to get event counts for issue",
-            extra={
-                "organization_slug": organization.slug,
-                "project_id": project_id,
-                "issue_id": issue_short_id,
-            },
-        )
-        return None
 
-    return {"count()": {"data": resp.data["data"]}}, stats_period, interval
+    if data is None:
+        return None
+    return data, stats_period, interval
 
 
 def get_issue_details(