Breaking: Grafana Cloud Now Accessible Directly from Terminal and AI Agents

In a move to close the visibility gap between coding agents and production environments, Grafana Labs today announced the public preview of gcx, a new command-line interface for Grafana Cloud. The tool enables engineers and their AI agents to access observability data—metrics, logs, traces, alerts, and SLOs—without leaving the terminal.

“The way developers work is shifting rapidly, with AI agents like Cursor and Claude Code handling more of the daily engineering tasks. But those agents have been blind to the live system,” said Jane Chen, VP of Product at Grafana Labs. “gcx gives them—and the humans supervising them—real-time production context directly in the command line, turning hours of context switching into seconds.”

From Zero to Fully Observable in Minutes

gcx is designed to handle the entire observability lifecycle from a single command-line session. Engineers can ask their agent to instrument a service, push alert rules, define SLOs, and verify data flow—all without opening a web browser. “Most services start with no instrumentation,” noted Chen. “gcx treats that as a starting line, not a blocker.”

Key capabilities available via the CLI include:

• Instrumentation: Wire OpenTelemetry into code and validate that metrics, logs, and traces are landing in the correct Grafana backends.
• Alerting & SLOs: Generate alert rules from actual service signals, define SLOs against real latency or availability data, and push them live.
• Synthetic monitoring: Set up synthetic probes so teams detect outages before users do.
• Frontend, backend, and Kubernetes monitoring: Onboard Faro-instrumented frontends, manage sourcemaps, and deploy Instrumentation Hub for backend services and Kubernetes infrastructure.
• Everything as code: Pull dashboards, alerts, SLOs, and checks as editable files, modify them locally with an agent, and push changes back—with a deep link to Grafana Cloud when a human needs to investigate further.

The Visibility Problem Agents Face

Agents that write code based solely on source files are pattern-matching without production context. They cannot see a latency spike during checkout or know whether SLOs are being met. “Without gcx, an agent is guessing,” said Alex Rivera, Senior Engineer at Grafana Labs. “With gcx, it reads the actual state of the running system and makes informed decisions based on what is happening, not what could happen.”

The tool eliminates the friction of switching from an agentic coding session to a separate observability dashboard. “What used to be a multi-day ticket becomes a one-agent session,” Rivera added.

Background

The rise of agentic coding assistants—such as Cursor and Claude Code—has dramatically accelerated code generation but created a new gap: these agents operate on local source code while remaining blind to production environments. Engineers spend hours context-switching between the terminal and observability tools to diagnose issues. gcx directly addresses this by embedding Grafana Cloud’s full observability stack into the command-line workflow used by both humans and agents.

Grafana Labs has long offered a rich web-based dashboard experience, but the company recognized that the future of engineering workflows is increasingly terminal-centric. The gcx CLI is the first tool from the company to treat the command line as a first-class interface for observability, not just a secondary access point.

What This Means

For engineering teams, gcx reduces the delay between writing code and understanding its production impact. By giving agents direct access to live telemetry, the tool enables more accurate code suggestions, faster incident response, and a tighter feedback loop. “This is a paradigm shift,” said Chen. “Observability should not be a separate step—it should be an integrated part of how you build and run software.”

The public preview is available now. Users can install gcx from the Grafana Cloud documentation or via package managers. Grafana Labs plans to extend the CLI with additional agent integrations and deeper workflow automation based on community feedback.

For more information, see the official gcx documentation.