--- title: "Grafana Integration" description: "Manage Grafana dashboards, folders, and data sources, record annotations, query metrics, and inspect alert rules from your flows over the Grafana HTTP API. Works with Grafana Cloud and self-hosted." url: https://flowrunner.ai/integrations/grafana date_modified: 2026-08-01T02:40:32-07:00 --- # Grafana [Developer Tools](https://flowrunner.ai/integrations/category/developer-infrastructure) Manage Grafana dashboards, folders, and data sources, record annotations, query metrics, and inspect alert rules from your flows over the Grafana HTTP API. Works with Grafana Cloud and self-hosted. [Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 23 actions · API key · available [Grafana website](https://grafana.com/) · [Platform Documentation](https://grafana.com/docs/grafana/latest/developers/http_api/) · Capability data verified 2026-07-14 1. A deploy flow finishes shipping a release 2. Agent resolves the dashboard the release affects 3. Agent calls Create Annotation to mark the release on the time-series graph 4. Agent posts the annotation and a link to the dashboard in the team channel 5. Agent runs Health Check to confirm the instance is reachable 6. An engineer confirms before the agent deletes any dashboard or folder during cleanup ## What This Integration Enables FlowRunner agents use Grafana to keep observability in step with the rest of the pipeline. An agent can create, update, and delete dashboards and folders as part of provisioning or CI/CD, record annotations to mark deploys, incidents, and events on time-series graphs, and query data sources directly to feed metrics into automations or alert conditions. It can audit alert rules, contact points, and organization users, and run a health check for uptime monitoring. What it does not do on its own is delete a dashboard or folder, because a delete removes the panels under it. That confirmation stays with a person. ### Without FlowRunner **Unmarked deploys**: Nobody can tell which metric spike lines up with which release **Manual dashboards**: A dashboard is built and updated by hand in the UI **Risky cleanup**: Someone deletes an old dashboard and loses panels they still needed ### With FlowRunner **Annotated releases**: The agent marks each deploy on the graph as it ships **Provisioned dashboards**: The agent creates and updates dashboards from CI/CD **Confirmed deletes**: Dashboard and folder deletes wait for an engineer to confirm ## Use Case Scenarios ### Deploy annotations for correlation A deploy flow finishes. The agent calls Create Annotation to record the release on the relevant dashboard, then posts the annotation and a dashboard link to the team channel so anyone chasing a metric change can jump straight to the graph. When an incident later opens, a second annotation marks its start time, and the two line up on the same timeline. Metric spikes become traceable to the change that caused them. ### Dashboards provisioned from CI/CD A new service ships with a dashboard defined in its repository. The agent calls Create or Update Dashboard to provision it under the right folder and points it at the correct data source. The service arrives with its dashboard in place, without anyone building panels by hand. ### Health check into a page A scheduled flow calls Health Check. When it reports the instance is unhealthy, the agent opens an incident in the on-call tool so a responder sees the outage. Monitoring the monitor becomes part of the same flow. ## Human-in-Loop Highlight The human-in-loop moment in Grafana is the deletion gate. Creating dashboards, recording annotations, querying data sources, and running health checks are all safe or reversible, so the agent runs them by rule. Deleting a dashboard or a folder is not, because it removes the panels and, for a folder, the dashboards inside it. When a cleanup flow flags dashboards that look stale, the agent lists them with their last-updated dates and asks in the observability channel: "These four dashboards have not been viewed in 90 days. Confirm which to delete?" An engineer confirms the list. Only then does the agent call Delete Dashboard. The agent keeps observability current on its own. A person owns the deletes that cannot be undone. Agent processes routinely Detects exception requiring judgment Clear match Continues automatically Ambiguous Routes to human via preferred channel Human decides Agent resumes with decision ## Agent Capabilities 23 actions ### Dashboards 5 - **Search Dashboards** Search dashboards by title, tag, or folder. - **Get Dashboard by UID** Retrieve a dashboard by its UID. - **Create or Update Dashboard** Create a new dashboard or update an existing one. Used to provision dashboards from CI/CD. - **Delete Dashboard** Delete a dashboard by UID. Run it behind a human confirmation. - **Get Home Dashboard** Retrieve the configured home dashboard. ### Folders 4 - **List Folders** List the dashboard folders. - **Get Folder** Retrieve a single folder by UID. - **Create Folder** Create a dashboard folder. - **Delete Folder** Delete a folder and, depending on settings, its dashboards. Run it behind a human confirmation. ### Data Sources 4 - **List Data Sources** List the configured data sources. - **Get Data Source** Retrieve a single data source. - **Create Data Source** Add a data source to the instance. - **Query Data Source** Run a query against a data source. Used to feed metrics into an automation or alert condition. ### Annotations 3 - **Create Annotation** Record an annotation to mark a deploy, incident, or event on a time-series graph. - **List Annotations** List annotations, filtered by time range, dashboard, or tag. - **Delete Annotation** Delete an annotation by ID. ### Alerting 3 - **List Alert Rules** List the alert rules configured in the instance. Used to audit alerting. - **Get Alert Rule** Retrieve a single alert rule. - **List Contact Points** List the alerting contact points. ### Organization & Users 4 - **Get Organization** Retrieve the current organization. - **List Organization Users** List the users in the organization. - **Get Current User** Retrieve the authenticated user. - **Health Check** Check the instance health. Used for uptime monitoring. ## Frequently Asked Questions ### What can FlowRunner do with Grafana? FlowRunner agents can run Search Dashboards, Get Dashboard by UID, and Create or Update Dashboard in Grafana, plus 20 more actions. ### Does connecting Grafana to FlowRunner require OAuth? No. Grafana connects to FlowRunner with an API key, no OAuth flow required. ### Can Grafana trigger a FlowRunner workflow automatically? Grafana doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at Grafana?** This integration exposes Grafana to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current. --- Markdown version of https://flowrunner.ai/integrations/grafana. Site index: https://flowrunner.ai/llms.txt