Adds the `usage_count` data source to query for a single Value Streams
Dashboard count metric and
return data fit for `SingleStat` visualization. Also creates
visualization templates for each metric.

Changelog: added
EE: true

This change updates the logic of
upserting and deleting policy rule by rule_index
when a security policy is updated.

EE: true
Changelog: fixed

Changelog: added
EE: true

This reverts merge request !183259

In addition, we unify all the usage in a single place.

Changelog: fixed

Given that configurable statuses functionality
should only be available for Premium and Ultimate
users, we need to ensure the status widget is
EE-only.

This commit moves the status widget-related classes
under the EE namespace to achieve that.

We do not need to support multiple values for filtering so we restrict
the GraphQL API argument.

This field is experimental and not yet used by any frontend query so we
can change it without deprecating.

Changelog: fixed
EE: true

Changelog: changed
EE: true

- the request for searchProjectMembers is causing issues
  in the tests, so we are going to disable them

EE: true

We moved the controls to a separate table and want to update the FE to
use the data from the new table.

Previously, the file size check required a complete graph walk,
which was slow.
This change optimizes the process by using an existing method
that checks only the quarantine directory, improving performance.

Additionally, since this is an EE feature,
a new EE-specific FileSizeCheck class was introduced
to retrieve and apply the push rule file size limit

Changelog: changed
EE: true

Currently, duo_chat and code_suggestions are enabled
Additionally, it includes troubleshoot job, explain vulnerability and
resolve vulnerability

Changelog: added
MR: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/176952
EE: true

To mimic Projects::BranchRules::MergeRequestApprovalSetting
for AllBranchesRules

In order to determine if there's any project approaching
the storage limit we need to query the database
instead of calculating on the fly

So we can delete top level groups without reloading the page

Contributes to https://gitlab.com/gitlab-org/gitlab/-/issues/523453

**Problem**

The 'Selective CodeOwners removal' configuration depends on the
approvals configuration at the group level. However, currently these
settings operate independently, which leads to an inconsistent
configuration state.

**Solution**

Disable "Selective CodeOwners removal" state if "Remove all approvals"
is enabled on group level.

Changelog: fixed
EE: true

This enables the feature flag by default
to rollout the severity override feature

Changelog: changed
EE: true

Changelog: fixed
EE: true

Changelog: fixed
EE: true

- Behind feature flag `vscode_extension_marketplace_settings`
- https://gitlab.com/gitlab-org/gitlab/-/issues/508977

On success, we should delete exports after 7 days. However, the
segmented export was overlooked and is still deleting exports after 1
hour.

Changelog: fixed
EE: true

We are moving away from the openbao proxy.
We will use JWT auth in the secrets manager client.

We need to track the various metrics related to
feature usage for Secrets Manager. One of the
metrics is the generation of ID token for accessing the secrets manager.

We define the event and required metrics to track
the ID token generation. We trigger the event when
the ID token is generated in the ProjectSecretsManager.

[Issue](https://gitlab.com/gitlab-org/gitlab/-/issues/428871)

Use `scripts/internal_events/cli.rb` to generate an event and a metric
Track the metric on #show in the controller
Add test to see that the metric was tracked

Instead of modifying the existing FeatureSettingsTable component, this commit adds new components that will
render the new expanded chat sub-features design. This will allow us to keep the new UI behind a feature flag so
there is more control over its release. Once expanded chat sub-features are stable we can remove the feature flag
and the old FeatureSettingsTable

EE: True

Ensure original and new index counts are captured at the same time after
indexing has paused. This prevents false positives when comparing document
counts caused by pending indexing operations.

EE: true