Application observability startup groundcover Ltd. today announced a major expansion of Agent Mode that lets artificial intelligence agents act on a team’s observability data across the development tools they already use.

The update allows engineers to direct agents to recommend code, open pull requests and manage tasks inside applications such as Slack, Linear and GitHub. The agent’s reasoning and execution never leave the customer’s own cloud. Each action ties back to a specific authorized user and runs under that user’s permissions.

Groundcover describes itself as a bring-your-own-cloud observability platform for the eBPF and OpenTelemetry protocols. Its pitch for the release is that the control teams keep by holding their data in their own environment is exactly what makes AI agents worth deploying. Most AI observability tools force a tradeoff, the company argues, because the agents work only if customers send sensitive, high-volume telemetry back to the vendor.

The expansion centers on new connectors that pull external applications, including Anthropic PBC’s Claude, Slack, Linear, GitHub and Cursor into the platform. Agent Mode can act within those tools using a user’s own credentials, while remote Model Context Protocol connectors let outside agentic services interface directly with Agent Mode. Because the agent runs inside the customer environment on full telemetry rather than a sampled slice, groundcover argues its recommendations reflect what is actually running in production.

A second addition lets organizations customize Agent Mode’s built-in skills or author their own, mapping the agent to the runbooks, operational playbooks and internal knowledge a team already follows. The release also adds administrator-level guardrails through centralized MCP authorization, allowing administrators to define which MCP services, connectors and tools the organization can use. Every execution and tool call is attributable to a named user.

“Our customers relied on groundcover dashboards to tell them what was happening in their stack, but that was never connected to the context sitting in applications such as Slack or Linear about why,” said groundcover Field Chief Technology Officer Noam Levy. “Now Agent Mode can work with both. Engineers can stay in the tools they already use, and the agent brings full telemetry context with it.”

Co-founder and Chief Technology Officer Yechezel Rabinovic added that keeping the agent where data lives means there is no token markup on top and said groundcover’s longer-term aim is to become the coordination layer between the engineer, the agent and the telemetry they depend on.

The features are available now at no additional cost and ship automatically to groundcover’s more than 200 deployed customers. The company plans to demonstrate the release during a launch webinar and an in-person event at its San Francisco meetup.

Image: groundcover