Grafana 13.1 Released: Observability as Code and a Wider AI Assistant

On June 24, 2026, Grafana Labs released Grafana 13.1, building on Grafana 13 by expanding observability as code, bringing Grafana Assistant to more data sources, and streamlining the everyday workflows teams use to visualize, analyze, and act on their data.

Git Sync: managing dashboards as code

Git Sync, which reached general availability with Grafana 13, brings native GitOps workflows into Grafana, with GitHub App authentication plus support for GitLab, BitBucket, and pure Git. Grafana 13.1 adds four more enhancements:

  • Import dashboards from the UI: You can now import dashboard JSON straight into a Git Sync-provisioned folder, picking the file path, branch, commit message, and workflow; the dashboard is committed back to your repository as part of the import. Uniqueness is path-based, and a conflicting path stops the import before anything is overwritten.
  • Root-level sync: You can sync dashboards at the root level without a containing folder, so provisioned dashboards sit alongside your non-provisioned ones — useful when a repo represents your whole Grafana setup.
  • Dashboard context visible by default: Git Sync-provisioned folders render their README.md inline by default, so ownership notes, runbooks, and links travel with the folder.
  • Automatically signed commits: Git Sync can now sign commits with GPG, SSH, or S/MIME keys so your Git provider marks them as verified — removing friction in repositories with branch protection that requires signed commits.

A typical repository structure looks like this:

dashboards/
├── README.md              # team conventions, naming rules
├── web/
│   ├── nginx-overview.json
│   └── README.md          # notes shown inside Grafana
└── database/
    └── mysql-performance.json

Combined with Codeowners review and grafana-toolkit validation in CI, dashboard changes get reviewed, reverted, and audited just like code. That is the real value of "dashboard as code": monitoring configuration stops being scattered across each person's browser.

Grafana Assistant reaches more data sources

Grafana Assistant is the AI-powered agent in Grafana Cloud. With 13.1 you can use it to directly query eight additional data sources — Snowflake, Jira, Dynatrace, and more — bringing the total to 30+. An investigation that starts with an alert can pull in error rates from Dynatrace, query performance from Oracle, and recent deployments from Jira all in one conversation, with no context switching. For each data source, Assistant queries with natural language, correlates signals across sources, and visualizes results as Grafana dashboards.

A concrete example: you get an alert that payment-service P95 latency is elevated. Ask Assistant "how does payment-service latency over the last hour relate to the most recent deployment?" It checks the latency curve in Dynatrace, finds the matching deployment in Jira, and compares error rates in your log source, then charts the correlation. A conclusion that once took four or five tool switches to assemble now starts with a single question.

Grafana Assistant is now also pre-installed in Grafana Enterprise with no plugin installation required; OSS users can still install the plugin from the catalog and connect their Grafana Cloud account.

Faster, more flexible dashboarding

  • Section-level variables are generally available: Previously, variables applied to the whole dashboard at once, which was a real limitation when a single dashboard spanned more than one service. Now each row or tab can carry its own independent variables, so an API gateway row can scope to one set of instances while a database row scopes to another, all in the same dashboard.

    "panels": [
  { "id": 1, "title": "API Gateway", "scopedVars": { "service": { "text": "gateway", "value": "gateway" } } },
  { "id": 2, "title": "Database", "scopedVars": { "service": { "text": "mysql", "value": "mysql" } } }
]

  • Revamped query editor in public preview: Multi-select with bulk actions lets you delete, hide, or show several queries at once, switch data sources in a single step, or enable/disable transformations in bulk; a stacked view lays out all queries, expressions, and transformations in one scrollable list.

  • More visualization updates: The Filter and Group by dashboard control is GA; a series visibility filter narrows series in time series legends; nested tables get cell styling and aggregation improvements; panel styles can be copied between panels, and curated style presets apply to time series, stat, gauge, bar gauge, and bar chart panels with one click.

PDC support for more data sources

Private Data Source Connect (PDC) creates a private, encrypted tunnel between your Grafana Cloud stack and data sources inside private networks, VPCs, or on-premises environments. Grafana 13.1 adds PDC support for three new data sources: MQTT (real-time IoT and sensor data), GitHub (GitHub Enterprise Server source control and project metrics), and IBM Db2 (on-premises or private cloud databases). Deploy a PDC agent inside your private network and configure the data source using its internal DNS name.

Upgrade and getting-started notes

13.1 is a minor upgrade within the 13.x line, so the risk is low. Before upgrading: back up grafana.db and custom configuration; check third-party plugin version compatibility; and run your Git Sync repository in a test environment before switching production. To get started, try the Prometheus + Grafana basics, reference the monitoring and alerting guide and website monitoring tools; smaller sites can also deploy Uptime Kuma. See more in the Monitoring & Alerting category.

16IDC Take

Two signals from Grafana 13.1 are worth noting: "dashboard as code" is becoming a collaboration baseline, with dashboards going through Git review and audit; and AI assistants are starting to genuinely bridge data sources, turning troubleshooting from tool-hopping into a single question.

FAQ

  • Can Git Sync coexist with traditional provisioning? Yes — root-level sync lets provisioned and non-provisioned dashboards sit together; migrate incrementally.
  • Do section-level variables break existing panels? No — default behavior is unchanged; only blocks that explicitly configure scoped variables use their own scope.
  • Does Assistant charge extra for the data sources it queries? It depends on the subscription and the source; if cost-sensitive, start with sources covered by your free allowance.

Source: https://grafana.com/blog/grafana-13-1-release-all-the-latest-features
Reference: Git Sync documentation https://grafana.com/docs/grafana/latest/developers/apis/dashboard-api/