Connect the Grafana MCP server to Unblocked to give AI agents live metrics, logs, and dashboard context

When a panel turns red, a developer has to find the right dashboard, rebuild the PromQL or LogQL behind it, and then guess which change moved the line. An agent working in the editor sees none of that. The Grafana Cloud MCP server exposes dashboards, data sources, metrics, logs, and incidents to AI agents, and Unblocked connects to it once so every agent can ask about your stack without its own Grafana setup.

This is a third-party MCP server: Grafana Labs hosts and maintains it, and Unblocked connects to it as a verified MCP server. Nothing is ingested or indexed — Unblocked runs the read-only queries you allow against your Grafana Cloud stack when a question needs them, so a latency graph or log stream reflects your systems at that moment, not a copy taken earlier.

Wiring Grafana’s MCP server straight into an agent gives it query results without the story behind them. Unblocked’s Context Engine reasons across those results and your other sources — tracing a latency jump on a dashboard panel to the deploy that preceded it, the pull request that changed the query path, and the Slack thread where the regression was first noticed — and returns one cited answer.

Once connected, any MCP-compatible coding agent — Claude Code, Cursor, Devin, GitHub Copilot, Codex — reaches Grafana context through a single Unblocked MCP server connection.

What you get

• Live metrics, logs, and dashboards — Agents run Prometheus and Loki queries, read dashboard summaries, and list incidents against your Grafana Cloud stack as it is now, scoped by your Grafana permissions.
• One connection for every agent — Authorize Grafana Cloud in Unblocked once. Claude Code, Cursor, Codex, and GitHub Copilot reach it through the Unblocked MCP server, with no per-agent Grafana configuration.
• Read-only, tool-level control — Tools that update dashboards, create incidents, or add annotations are never offered. You pick which read-only Grafana tools Unblocked may call from Settings → Data Sources.
• Context Engine reasoning — Grafana results are reconciled with your code, pull requests, tickets, and conversations, so a spike on a panel leads to the change behind it — with citations.

Prerequisites

• An Unblocked account (free to start).
• A Grafana Cloud stack with Grafana Assistant available, and the Assistant Cloud MCP User role (Editors and above have it by default) to authorize the OAuth connection.
• An MCP-compatible coding agent (Claude Code, Cursor, Devin, etc.) to use the context in your editor.

Step 1: Connect Grafana to Unblocked

1. In Unblocked, go to Settings → Data Sources → Connect another data source, and under Incident Management select Grafana.
2. The Grafana MCP server URL is filled in for you. Choose OAuth, click Connect, and sign in to Grafana to authorize Unblocked.
3. Once Unblocked discovers the server’s tools, select the read-only tools it may use and click Save Settings.

Step 2: Install the Unblocked MCP server

On macOS or Linux, one command auto-detects and configures the supported agents on your machine:

On Windows, or when you prefer remote MCP, add the server manually using the configuration below, or follow the manual install guide .

claude mcp add -s user --transport http unblocked https://getunblocked.com/api/mcpsse

{
  "mcpServers": {
    "unblocked": {
      "url": "https://getunblocked.com/api/mcpsse"
    }
  }
}

https://getunblocked.com/api/mcpsse

Grafana hosts and maintains this MCP server. Unblocked calls only the read-only tools you select, at the moment a question needs them:

• search_dashboardsSearches dashboards in your Grafana Cloud stack and returns the matching dashboards.
• get_dashboard_summaryReturns a compact overview of a dashboard — its panels and their queries — without the full dashboard JSON.
• query_prometheusRuns a PromQL query against a Prometheus data source and returns the resulting series.
• query_loki_logsRuns a LogQL query against a Loki data source and returns the matching log lines.
• list_incidentsLists incidents recorded in Grafana Incident, with their status and details.

To give Claude Code, Cursor, GitHub Copilot, Codex, or another MCP-compatible AI coding agent access to Grafana context, connect Grafana to Unblocked and add the single Unblocked MCP server to your agent.

Unblocked acts as the MCP context layer for your Grafana data, searching metrics, logs, and dashboards alongside your other connected sources and returning cited results through these tools:

• context_researchResearches a question across your connected sources and calls the Grafana tools you selected when metrics, logs, dashboards, or incidents are relevant, returning a synthesized answer with citations.
• context_get_urlsFetches the full content behind URLs your agent encounters across connected sources — Jira, Linear, Notion, Confluence, Slack, GitHub, and more.

Questions you (or your agent) can ask Unblocked about your Grafana data:

• Why did tail latency on the checkout dashboard climb this morning, and which pull request merged right before it?
• What do the Loki logs for the payments service show since the last deploy, and has anyone raised it in Slack?
• Which Prometheus metrics does the orders service emit, and where in the code are they recorded?
• Summarize the open Grafana incidents and link each one to the ticket or PR that is fixing it.