groundcover has expanded Agent Mode with so-called Connectors, which allow the AI assistant not only to analyze observability data but also to perform actions in development and collaboration tools such as GitHub, Slack, Linear, Cursor, and Claude Desktop.
Agent Mode is the AI assistant that forms part of groundcover’s observability platform. The feature can answer questions in natural language about logs, traces, metrics, Kubernetes events, and other telemetry data. This enables the agent to investigate incidents, determine the cause of outages, or set up dashboards and monitoring rules.
With the new Connectors, the AI agent can also operate outside the observability platform. For example, engineers can have the agent suggest code, create pull requests, or manage tasks in external applications. The AI uses the relevant user’s access permissions, ensuring that only actions for which that person is authorized are performed.
Customer’s Own Cloud Environment
According to groundcover, Agent Mode continues to run within the customer’s own cloud environment—an architecture the company refers to as Bring Your Own Cloud (BYOC). As a result, telemetry data does not need to be transferred to the vendor’s infrastructure. The AI analyzes the data where it already resides.
In addition, groundcover introduces centralized management capabilities for organizations. Administrators can determine which external AI services, MCP integrations, and connectors are available. All actions performed by the AI agent are linked to a specific user and are subject to existing role-based access rights.
Customizable AI Agent
Another new feature is the ability to extend Agent Mode with custom capabilities. Organizations can equip the AI with internal runbooks, operational procedures, and other workflows, ensuring the agent aligns more closely with existing processes.
Through support for the Model Context Protocol (MCP), Agent Mode can also collaborate with external AI agents. With this, groundcover aims to make observability data directly available within the development toolchain, without requiring engineers to switch between different applications.
With this expansion, groundcover is aligning with a broader trend in the observability market. More and more vendors are combining AI with monitoring platforms, but groundcover is explicitly focusing on AI agents that not only perform analyses but can also take concrete actions within the day-to-day development environment. The new capabilities are automatically made available to the platform’s more than 200 customers.