Documentationbreadcrumb arrow Grafana Cloudbreadcrumb arrow AI and machine learningbreadcrumb arrow Grafana Assistantbreadcrumb arrow Guidesbreadcrumb arrow Navigate Grafana resources
Grafana Cloud

Navigate Grafana resources

Use the Grafana Assistant to locate dashboards, discover data sources, and navigate directly to troubleshooting views. Instead of browsing folders, you can describe what you need using natural language.

Before you begin

Ensure the Assistant has permission to access your data sources and dashboards. For more information, refer to Security and privacy.

Find dashboards and panels

Describe what you need using tags, metric names, or queries to find relevant visualizations.

Search by content

Ask the Assistant to find dashboards based on their content or tags.

Find dashboards tagged “production” that include checkout metrics.

You can also find specific panels by describing the data they visualize.

Locate panels that query rate(http_requests_total[5m]) and provide links.

Open resources with context

When you ask the Assistant to open a dashboard, it preserves your current time range, so you don’t lose context.

Open the dashboard with UID <DASHBOARD_UID>.

Compare dashboards

If multiple dashboards match your request, ask for a comparison to decide which one to use.

Compare the checkout and cart dashboards.

Discover data sources and metrics

Inventory your environment to understand available data sources and metrics.

List data sources

Filter data sources by type, name, or environment using @ mentions where applicable.

List Prometheus data sources available to me.

List Loki data sources with names containing “prod”.

Explore metrics

Drill into available metrics and labels without leaving the chat.

Search for metrics matching cart.* error rates in @prometheus-ds.

List label values for namespace in @prometheus-ds for the last six hours.

Format findings

Request specific output formats, such as lists or tables, to organize the information.

List Prometheus data sources as a markdown table.

Coordinate incident response

During an incident, use the Assistant to direct responders to the right views and tools quickly.

Gather relevant assets

Collect all dashboards and runbooks related to a specific service or problem.

Surface dashboards and runbooks relevant to the checkout latency incident.

Summarize which dashboards and tools we are using for the checkout incident and include links.

Launch tools with context

Direct the Assistant to open Explore with a specific query already loaded, saving setup time for responders.

Open Explore with a LogQL query that filters checkout-service errors in the last 30 minutes.

Next steps