Applicative metrics

Export application metrics to Prometheus

GitGuardian exposes application metrics in the Prometheus format, so you can scrape them for monitoring and alerting.

info

The Prometheus exporter is not supported for embedded clusters.

Metrics available

GitGuardian exports the following metrics, grouped by area.

Metric names in Prometheus

Metrics are listed with the names exposed in Prometheus. Some gauges also end in _total, so check the Type column before using rate(). Prometheus appends _bucket, _sum, and _count to histogram names, even when the name already ends in _bucket. For example, gim_celery_task_runtime_bucket is queried as gim_celery_task_runtime_bucket_bucket. On self-hosted instances, the plan label is always on_prem.

Where metrics are served

Each metric is served by the pod that records it:

• app-exporter service: gauges computed from the database and Redis (users, incidents, commits, active scans, infrastructure, queue length, installations, health checks, periodic tasks, and Public API quota, usage, and tokens).
• Celery worker pods (port 9808): metrics recorded by background tasks, which cover most counters and histograms on this page.
• Web application pods (/metrics on port 5050, or 5051 for public-api): metrics recorded while serving requests, such as HTTP requests, authentication, SCIM, Public API calls, and webhook events received by the hook pods.

The provided Service Monitors cover app-exporter, the Celery workers, and the ML Secret Engine when deployed. They do not cover web application pods, except hook pods when observability.exporters.webAppExporter.enabled is true. To collect metrics from the other web application pods, add your own ServiceMonitor or PodMonitor.

Each pod reports only its own values, so aggregate across pods in your queries.

Users and incidents

Metric	Type	Description	Dimensions
gim_active_users_total	Gauge	All users in the instance	None
gim_billable_users_total	Gauge	Contributing developers counted towards your license	None
gim_issues_total	Gauge	All incidents	severity, status
gim_occurrences_total	Gauge	All occurrences	hidden, status

Scanning activity

Metric	Type	Description	Dimensions
gim_commits_total	Gauge	Commits processed	account, scan_type
gim_repo_scan_active_statuses_total	Gauge	Scans currently pending or running, by scan feature	scan_feature, status, plan
gim_scan_ended_total	Counter	Scans that ended, successfully or not	framework_version, source_type, status, scan_type, plan, failure_reason, exception, fault_domain
gim_scan_pending_duration_seconds	Histogram	Time a scan spent pending before it launched	framework_version, source_type, status, scan_type, plan
gim_scan_duration_seconds	Histogram	Scan duration, from launch to end	framework_version, source_type, status, scan_type, plan
gim_scan_total_duration_seconds	Histogram	End-to-end scan duration, from creation to end	framework_version, source_type, status, scan_type, plan, trigger_origin, trigger_family
gim_scan_elements_scanned_total	Counter	Elements scanned	source_type, scan_method, plan
gim_scan_elements_skipped_total	Counter	Elements skipped during scans	source_type, reason, plan
gim_scan_elements_failed_total	Counter	Elements that failed during scans	source_type, reason, plan
gim_scan_data_downloaded_bytes_total	Counter	Bytes downloaded from sources during scans	source_type, plan
gim_scan_data_scanned_bytes_total	Counter	Bytes scanned for secrets during scans	source_type, plan
gim_historical_scan_failure_total	Counter	Deprecated, use gim_scan_ended_total. Scans that ended without success	integration_name, scan_feature, scan_status, exception, failure_reason
gim_historical_scan_duration_seconds_bucket	Histogram	Duration of repository scans and of historical or recurrent scans of other sources	scan_feature, plan
gim_vcs_scan_deferred_total	Counter	VCS scans deferred because the workspace reached its maximum number of simultaneous VCS scans	retry_number
gim_vcs_scan_retry_scheduled_total	Counter	VCS scan retries scheduled after a failure	source_type, failure_reason
gim_vcs_scan_retry_outcome_total	Counter	Final status of VCS scan retries	source_type, status
gim_vcs_clone_time_seconds_bucket	Histogram	Time taken to clone a repository	integration_name, plan, clone_method
gim_tokenscanner_content_oversized_total	Counter	Content not scanned because it exceeds the maximum scan size	caller

• In gim_repo_scan_active_statuses_total, the scan_feature="check_run_pr" series repeats scans already counted under secrets. Exclude it when summing across features.
• The gim_scan_pending_duration_seconds, gim_scan_elements_*, and gim_scan_data_* metrics only cover non-VCS sources (for example Slack, Jira, or Confluence).
• gim_historical_scan_failure_total counts failed, timed out, skipped, canceled, and too large scans. Use scan_status to tell them apart.
• The limit on simultaneous VCS scans is disabled by default, so gim_vcs_scan_deferred_total usually reports no data.

GitHub check runs

Metric	Type	Description	Dimensions
gim_check_runs_init_total	Counter	Check runs initialized, whether or not they end up being created	created, setting_from, plan
gim_check_runs_created_total	Counter	GitHub check runs created	plan, account_id
gim_check_runs_timed_out_total	Counter	GitHub check runs that timed out	plan, account_id
gim_check_runs_runtime_bucket	Histogram	GitHub check run duration, including queue time	status, plan
gim_check_runs_error_codes_total	Counter	GitHub check run errors by code	error_code
gim_check_runs_pr_too_large_skipped_total	Counter	Check runs completed as neutral without a scan, because the pull request exceeded a size threshold	plan, reason
gim_check_runs_pr_too_large_scan_dispatched_total	Counter	Pull requests too large for the regular check run path, scanned with a scan-based check run instead	plan, reason
gim_check_runs_softtimelimit_retry_skipped_total	Counter	Check runs not retried after hitting the task time limit	placeholder
gim_github_call_errors_total	Counter	GitHub errors when listing a branch's pull requests, counted on retried check runs	action, error_class

For the pr_too_large metrics, reason is the exceeded threshold: commits, files, or changes.

Real-time monitoring

Metric	Type	Description	Dimensions
gim_realtime_event_received_total	Counter	Real-time events received from integrations	integration_name, event_type
gim_realtime_event_evaluation_started_total	Counter	Real-time events whose evaluation started	integration_name, plan, event_type
gim_realtime_event_evaluation_success_total	Counter	Real-time events evaluated successfully	integration_name, plan, event_type
gim_realtime_event_evaluation_failure_total	Counter	Real-time events whose evaluation failed and will not be retried	integration_name, exception, plan, event_type
gim_realtime_event_retry_total	Counter	Real-time events retried after their evaluation hit the task time limit	integration_name, exception, plan, event_type
gim_realtime_event_dropped_total	Counter	Real-time events dropped	integration_name, plan, event_type
gim_realtime_elements_scanned_total	Counter	Real-time elements scanned for secrets	integration_name, plan
gim_realtime_elements_skipped_total	Counter	Real-time elements skipped before scanning	integration_name, plan, reason
gim_realtime_elements_failed_total	Counter	Real-time elements that failed during scanning	integration_name, plan, exception

Incidents and secrets

Metric	Type	Description	Dimensions
gim_issue_seen_total	Counter	Incidents seen in scans	policy_name, type, kind, detector_name
gim_occurrence_seen_total	Counter	Occurrences seen in scans	integration_name, policy_name, type, kind
gim_secret_validity_check_duration_seconds	Histogram	Duration of secret validity checks	detector_name, validity_status, has_custom_host, timed_out
gim_secret_validity_status_transition_total	Counter	Validity check results, by previous and new validity status, including unchanged statuses	detector_name, previous_status, new_status
gim_presence_check_duration_seconds	Histogram	Duration of presence checks	kind
gim_presence_check_success_total	Counter	Occurrences checked for presence	kind
gim_presence_check_updated_total	Counter	Occurrences updated by presence checks	kind, presence_value
gim_playbook_activated_total	Counter	Playbooks activated	playbook_type, is_public
gim_playbook_deactivated_total	Counter	Playbooks deactivated	playbook_type, is_public
gim_playbook_issues_resolved_total	Counter	Incidents resolved by playbooks	playbook_type, is_public
gim_playbook_issues_ignored_total	Counter	Incidents ignored by playbooks	playbook_type, is_public
gim_playbook_issues_unignored_total	Counter	Incidents unignored by playbooks	playbook_type, is_public

Infrastructure

Metric	Type	Description	Dimensions
gim_postgres_used_disk_bytes	Gauge	Size of the GitGuardian PostgreSQL database, excluding WAL files	None
gim_redis_used_memory_bytes	Gauge	Memory used by the Redis cache instance	None
gim_redis_available_memory_bytes	Gauge	Total memory of the host running Redis	None

gim_redis_available_memory_bytes is the Redis total_system_memory value, not maxmemory or a container memory limit. On Kubernetes, it is the memory of the node running Redis.

Celery workers and queues

Metric	Type	Description	Dimensions
gim_celery_queue_length	Gauge	Tasks waiting in queue	queue_name, team
gim_celery_active_consumer_count	Gauge	Worker processes consuming the queue, reported by each worker pod	queue_name, team
gim_celery_worker_up	Gauge	Running Celery workers, each reporting 1 once started	placeholder
gim_celery_worker_tasks_active	Gauge	Tasks currently being processed	placeholder
gim_celery_task_sent_total	Counter	Tasks sent to a queue	name, queue_name, team
gim_celery_task_received_total	Counter	Tasks received by a worker	name, queue_name, team
gim_celery_task_started_total	Counter	Tasks started by a worker	name, queue_name, team
gim_celery_task_succeeded_total	Counter	Tasks completed successfully	name, queue_name, team
gim_celery_task_failed_total	Counter	Tasks that failed	name, exception, queue_name, team
gim_celery_task_retried_total	Counter	Tasks retried	name, queue_name, team
gim_celery_task_revoked_total	Counter	Tasks revoked	name, queue_name, team
gim_celery_task_queue_time_bucket	Histogram	Time spent waiting in queue	name, queue_name, team
gim_celery_task_runtime_bucket	Histogram	Execution duration of tasks that completed successfully	name, queue_name, team

The placeholder label is always ok. Tasks sent while serving a request are counted in gim_celery_task_sent_total by the web application pods. Periodic tasks sent by the Celery beat scheduler are not counted.

HTTP requests

Metric	Type	Description	Dimensions
gim_http_request_started_total	Counter	HTTP requests initiated	api, method, view_name
gim_http_request_success_total	Counter	HTTP requests that returned a status below 400	api, method, view_name, plan, ggshield_command, business_account_id
gim_http_request_failure_total	Counter	HTTP requests that returned a 4xx or 5xx status	api, method, status_code, view_name, plan, ggshield_command, business_account_id
gim_http_request_exception_total	Counter	API requests where the view raised an unhandled exception	api, method, view_name, exception, plan
gim_http_request_view_duration_seconds_bucket	Histogram	Time spent in the view, excluding middleware	api, method, view_name, plan
gim_http_request_middleware_before_duration_seconds_bucket	Histogram	Time spent in middleware before the view	api
gim_http_request_middleware_after_duration_seconds_bucket	Histogram	Time spent in middleware after the view	api
gim_http_request_body_bytes_bucket	Histogram	Size of the request body, taken from the Content-Length header	api
gim_requests_with_invalid_host_header_total	Counter	Requests rejected because of an invalid Host header	api, method

Integrations and health checks

Metric	Type	Description	Dimensions
gim_installations_total	Gauge	Integration installations in any state	integration_type, account_id
gim_installations_status	Gauge	Integration installations by state	integration_type, account_id, state
gim_installation_uptime_seconds	Histogram	Time an installation stayed healthy before becoming unhealthy	integration_type, account_id
gim_installation_downtime_seconds	Histogram	Time an installation stayed unhealthy before recovering	integration_type, account_id
gim_integration_oauth2_access_tokens_expiring_soon	Gauge	Valid OAuth2 access tokens that expire within the renewal margin (Slack, Jira Cloud, and Confluence Cloud)	integration
gim_integration_api_call_success_total	Counter	Successful API calls to integrations	integration_name, method, status_code, api_endpoint
gim_integration_api_call_failure_total	Counter	Failed API calls to integrations	integration_name, method, status_code, exception, api_endpoint
gim_integration_api_call_retry_total	Counter	Retry attempts of API calls to integrations	integration_name, method, exception, api_endpoint
gim_integration_api_call_rate_limit_total	Counter	API calls to integrations that hit the provider's rate limit	integration_name, method, status_code, api_endpoint
gim_integration_api_call_duration_total	Counter	Cumulative duration of API calls to integrations, in seconds	integration_name, method, status_code, api_endpoint
gim_integration_health_check_probe_failure_total	Counter	Failed health check probes to integrations, counted separately from API call failures	integration_name, method, status_code, exception, api_endpoint
gim_integration_app_credential_invalid_total	Counter	App-level credentials of an integration rejected by the provider	integration_name, reason
gim_integrations_oauth2_renewal_total	Counter	OAuth2 token renewal attempts	integration, outcome, failure_reason, attempt_number
gim_integration_sync_runs_total	Counter	Installation synchronization runs	integration_name, sync_type, source_type, outcome
gim_health_check_result_count	Gauge	Services by result of their latest health check, refreshed daily	service_name, status
gim_outdated_health_check_count	Gauge	Health checks older than the periodic range	service_name
gim_health_check_skipped_total	Counter	Health checks skipped	reason, service_name
gim_health_check_duration_metric	Histogram	Health check execution duration	service_name

Some integration metrics are reported only by specific integrations:

• gim_integration_api_call_retry_total: Slack.
• gim_integration_api_call_rate_limit_total: GitLab, Microsoft Teams, SharePoint Online, OneDrive, Amazon ECR, and Amazon S3. All except GitLab also count these calls in gim_integration_api_call_failure_total, so do not add the two together.
• gim_integration_app_credential_invalid_total: Microsoft Teams, SharePoint Online, and OneDrive.

Notifications and emails

Metric	Type	Description	Dimensions
gim_notifier_run_failure_total	Counter	Notifier runs that failed	notifier_name, exception, source_integration_name, event_type
gim_ebn_send_notification_event_total	Counter	Notification send attempts, including retries, failures, and ignored notifications	event_type, ebn_service_name, is_retry
gim_ebn_send_notification_event_success_total	Counter	Notifications sent successfully	event_type, ebn_service_name, is_retry
gim_ebn_send_notification_event_failure_total	Counter	Notifications that failed to be sent	event_type, ebn_service_name, is_retry, retryable
gim_ebn_send_notification_event_ignored_total	Counter	Notifications ignored without being sent, for example when a new incident is already ignored or resolved	event_type, ebn_service_name, is_retry
gim_ebn_send_notification_event_duration_seconds	Histogram	Time taken to send a notification	event_type, ebn_service_name, is_retry
gim_email_sent_success_total	Counter	Emails sent without error, counted per recipient	template_name
gim_email_sent_failure_total	Counter	Emails that failed to be sent, counted per recipient	template_name, exception

Emails sent while serving a request, such as MFA codes, are counted by the web application pods. gim_email_sent_success_total also increases when no email server (SMTP or SendGrid) is configured, even though no email is sent.

Authentication and provisioning

Metric	Type	Description	Dimensions
gim_authentication_sso_method_total	Counter	SAML responses that passed signature and authentication checks, by authentication method reported by the identity provider	auth_method, idp_type
gim_saml_sso_flow_started_total	Counter	SAML SSO logins started	idp_type
gim_saml_sso_sca_received_total	Counter	SAML assertions received from the identity provider	idp_type
gim_saml_sso_sca_rejected_total	Counter	SAML assertions rejected	idp_type, error_type
gim_mfa_code_generated_total	Counter	MFA codes generated	flow, is_resend
gim_mfa_rate_limit_total	Counter	MFA requests rejected by rate limiting	flow, kind
gim_mfa_code_verification_total	Counter	MFA code verifications	flow, is_success
gim_scim_user_created_total	Counter	Users created through SCIM	account_id
gim_scim_saml_user_created_total	Counter	SAML users created through SCIM	account_id
gim_scim_member_created_total	Counter	Memberships created through SCIM	account_id
gim_scim_member_deleted_total	Counter	Memberships deleted through SCIM	account_id
gim_scim_error_total_metric_total	Counter	Errors returned by the SCIM API	status_code, scim_type, operation
gim_members_deactivated_total	Counter	Members deactivated	plan
gim_members_reactivated_total	Counter	Members reactivated	plan

Periodic tasks

Metric	Type	Description	Dimensions
gim_periodic_task_period_seconds	Gauge	Expected periodicity of a task	task_name, queue_name
gim_periodic_task_not_run_for_seconds	Gauge	Time since the scheduler last sent the task	task_name, queue_name

Public API

Metric	Type	Description	Dimensions
gim_public_api_quota_total	Gauge	Maximum allowed API usage	account
gim_public_api_usage_total	Gauge	Current API usage	account
gim_public_api_token_total	Gauge	Active API tokens	account, type
gim_public_api_scan_duration_seconds	Histogram	Duration of content scans through the API	is_business, preset_name, all_secrets, multiscan
gim_public_api_scan_total_documents_size_bytes	Histogram	Total size of documents scanned through the API	is_business, preset_name, multiscan
gim_invalid_api_key_total	Counter	API calls rejected because of an invalid API key	method, path, api_type, view_name, reason, token_id
gim_quota_limit_reached_total	Counter	API calls rejected because the quota was reached	plan, business_account_id
gim_ggshield_calls_total	Counter	Calls made by ggshield	ggshield_version, ggshield_os_name, ggshield_python_version, ggshield_command

Recommended monitoring

System health

Metric	What to watch for
gim_celery_queue_length	A backlog that keeps growing is an early sign of saturation.
gim_celery_active_consumer_count	A drop means worker pods stopped. A queue with no consumer has no series at all: see the example queries below.
gim_postgres_used_disk_bytes	Track database growth. Also monitor the PostgreSQL volume disk usage.
gim_redis_used_memory_bytes / gim_redis_available_memory_bytes	A high ratio indicates memory pressure on the Redis host. To detect Redis reaching its own limit, compare with maxmemory or the pod memory limit.
gim_periodic_task_not_run_for_seconds	Compare against gim_periodic_task_period_seconds to detect stalled periodic tasks.

API and integrations

Metric	What to watch for
gim_http_request_failure_total	Filter by status_code to spot error rate spikes (especially 5xx).
gim_http_request_view_duration_seconds_bucket	Track view latency (P95/P99) to detect slowdowns. It excludes middleware time.
gim_health_check_result_count	Filter by status="not_ok" to detect broken integrations. Refreshed daily, so it can lag by up to a day.
gim_integration_api_call_failure_total	Filter by integration_name and status_code to spot a provider rejecting or failing API calls.
gim_check_runs_timed_out_total	A rising rate means GitHub check runs are not completing in time.
gim_check_runs_pr_too_large_skipped_total	A rising rate means more pull requests are skipped for size.
gim_scan_ended_total	Filter by status!="finished" to spot unsuccessful scans, including canceled ones. fault_domain tells where the failure comes from: gim (GitGuardian), customer (access, deleted resources, rate limits, or size limits on your side), or provider (temporary source provider errors). Alert on fault_domain="gim".
gim_installations_status	Compare with gim_installations_total to track installations by state.
gim_ebn_send_notification_event_failure_total	Failures with retryable="False" are final. A sustained rate means notifications are not delivered.

HTTP request metrics are served by the web application pods, which the provided Service Monitors do not cover (see Where metrics are served).

Example PromQL queries

# Celery queue saturation (waiting tasks per active consumer)
max by (queue_name) (gim_celery_queue_length) / sum by (queue_name) (gim_celery_active_consumer_count)

# Queues with waiting tasks and no active consumer (not shown by the saturation query above)
max by (queue_name) (gim_celery_queue_length) > 0 unless on (queue_name) sum by (queue_name) (gim_celery_active_consumer_count)

# HTTP 5xx error rate over the last 5 minutes
sum(rate(gim_http_request_failure_total{status_code=~"5.."}[5m]))

# View latency P95 over the last 5 minutes (excludes middleware time)
histogram_quantile(0.95, sum by (le) (rate(gim_http_request_view_duration_seconds_bucket_bucket[5m])))

# Redis memory as a share of the host's total memory (%)
gim_redis_used_memory_bytes / (gim_redis_available_memory_bytes > 0) * 100

# Periodic tasks the scheduler has not sent for 2x their expected period
gim_periodic_task_not_run_for_seconds > 2 * gim_periodic_task_period_seconds

Example alerting rules

Below are starter Prometheus alerting rules you can adapt to your environment:

groups:
  - name: gitguardian
    rules:
      - alert: CeleryQueueBacklog
        expr: max by (queue_name) (gim_celery_queue_length) > 100
        for: 5m
        annotations:
          summary: "Queue {{ $labels.queue_name }} has {{ $value }} pending tasks"

      - alert: RedisMemoryPressure
        expr: gim_redis_used_memory_bytes / (gim_redis_available_memory_bytes > 0) > 0.85
        for: 10m

      - alert: PeriodicTaskStalled
        expr: gim_periodic_task_not_run_for_seconds > 2 * gim_periodic_task_period_seconds
        for: 5m

      # Requires scraping the web application pods
      - alert: HighHTTPErrorRate
        expr: sum(rate(gim_http_request_failure_total{status_code=~"5.."}[5m])) > 0.1
        for: 5m

If Redis has maxmemory set or a container memory limit, adjust the RedisMemoryPressure threshold.

Visualizing metrics with Grafana

GitGuardian does not ship pre-built Grafana dashboards, but all gim_* metrics are standard Prometheus metrics and work with any Grafana instance connected to your Prometheus server.

To get started:

1. Add your Prometheus server as a Grafana data source.
2. Create panels using the PromQL queries above.
3. Organize panels into rows: System health (queues, Redis, Postgres, periodic tasks) and API and integrations (error rates, health checks).

Enable or disable application metrics

Celery worker and web application pods expose their metrics without extra configuration. To activate the app-exporter gauges:

• authorize collection of metrics by the application
• activate Prometheus export

Authorize metrics collection

In the Admin Area, go to Preferences, check the prometheus_metrics_active feature flag, and save.

To disable it, uncheck the flag and save.

While unchecked, the app-exporter service returns a 404 error. Celery worker and web application metrics are not affected.

Install Prometheus Operator

Metrics are collected by Prometheus using the Prometheus Operator.

On Existing Clusters, install it yourself (installation documentation).

Activate Prometheus export using KOTS

In the KOTS Admin Console, open the configuration page and check Activate Prometheus Exporter in the Prometheus section.

Save the configuration and Deploy the application. This deploys the app-exporter service and creates the Service Monitors.

To disable it, uncheck the box, save, and deploy again.

Activate Prometheus export using Helm

To deploy the app-exporter service, set observability.exporters.statefulAppExporter.enabled=true in your values file.

observability:
  exporters:
    statefulAppExporter:
      enabled: true

If you use Prometheus Operator, set observability.serviceMonitors.enabled=true to create the Service Monitors. Set observability.exporters.webAppExporter.enabled=true to also cover the hook pods.

observability:
  exporters:
    webAppExporter:
      enabled: true
    statefulAppExporter:
      enabled: true
  serviceMonitors:
    enabled: true

Otherwise, you can manually scrape the metric endpoints:

• app-exporter: http://app-exporter.<namespace>.svc.cluster.local:9808/metrics
• Celery workers: each pod on port 9808, at /metrics. Use Kubernetes endpoints discovery to scrape every pod, since the worker service (for example worker or scanners) only reaches one pod per request.
• Web application pods: each pod at /metrics, on port 5050 (5051 for public-api).

How to collect metrics

On Existing Clusters, you install and configure Prometheus yourself. If you use the kube-prometheus stack with Service Monitors enabled, Prometheus discovers the covered pods automatically (see Where metrics are served).

Your Prometheus instance must select these Service Monitors. For example, kube-prometheus-stack only selects Service Monitors that carry its release label. With Helm, set it with observability.serviceMonitors.labels (see Autoscaling).

Otherwise, scrape the endpoints listed in Activate Prometheus export using Helm.

Usage data

GitGuardian collects usage data to improve the product and support. To turn it off, uncheck custom_telemetry_active in the preferences section of the Admin area.

info

Why keep usage data enabled?

• Product improvement: usage data shows us how the application runs across environments, so we know where to focus fixes and testing.

• Faster support: when something breaks, usage data helps us find and fix the issue faster, which means less downtime for you.

• Privacy: we do not collect personal or sensitive data.

Here are the categories and metrics we collect:

• Replicated

  • Various deployment-related metrics such as cloud status, version, and uptime

• System

  • SSO Provider
  • Network Gateway Type
  • Custom CA and proxy status
  • Prometheus Application metrics activation status

• Users & Teams

  • Number of pending invitees, registered users with different access levels, active users.

• Billing

  • Number of contributing developers counted towards your license, and the number excluded from the count
  • The same two counts broken down per VCS type (Azure DevOps, Bitbucket Cloud, Bitbucket Data Center, Gerrit, GitHub, GitHub Enterprise, GitLab)

  These are aggregate counts of distinct commit author email addresses. The email addresses themselves are never transmitted.

• Historical Scan

  • Number of historical scans canceled, failed, and finished
  • Percentile durations of historical scans
  • Number of secrets found per day and source scans per day in historical scans
  • Number of historical scans considered too large

• Integrations

  • Number of instances, installations, projects, sites for VCS and Other Data Sources
  • Number of monitored and unmonitored sources, along with source size percentiles and estimated users per VCS

• Secret

  • Number of deactivated detectors per category, registered and unregistered feedbacks, and various metrics related to secret validity checks and incidents

• Public API

  • Number of calls for ggshield secrets scans, including different modes and repositories
  • Number of active personal access tokens and service accounts
  • Number of public API calls