Application observability startup groundcover Inc. today gave outside artificial intelligence coding agents such as Anthropic PBC’s Claude and OpenAI Group PBC’s Codex a way to hand production investigations to its own observability agent.

The launch, which groundcover is calling agent-to-agent observability, is seeking to address what happens after code ships. Coding agents can already review a pull request and work out from the code how software is meant to behave, but their view usually ends at deployment.

Shahar Azulay, groundcover’s co-founder and chief executive, said agents “can’t tell you what already did happen in production unless they have access to the evidence.” An agent can now delegate the digging to groundcover, he said, without having to “become an observability expert.”

Outside agents get two ways in. An agent that already knows which evidence it wants can pull logs, traces and Kubernetes events straight from groundcover’s Model Context Protocol tools. Vaguer problems go to the groundcover Agent as asynchronous Agent Tasks, and a job ID comes back right away. The investigation then runs in the background while the calling agent moves on, and the findings arrive later as a completed root cause analysis or a draft dashboard, depending on what was asked for.

Cursor and Amazon Web Services Inc.’s Kiro can connect through MCP alongside Claude and Codex. The launch also reaches two places outside the coding tools. Inside groundcover itself, Agent Mode lets engineers steer an investigation in real time or have the agent generate dashboards and monitors from what it finds.

In Slack, teams can bring @groundcover into an incident thread, where it treats the conversation as context and can draw on connected systems such as Linear and GitHub. Groundcover first connected Agent Mode to Slack, Linear and GitHub in June, and the incident-thread feature builds on that.

Every part of the launch runs on groundcover’s bring-your-own-cloud architecture, so telemetry stays inside the customer’s own cloud account. A single sensor captures application and infrastructure behavior without every service having to be instrumented. Customers pay by the node.

Because data volume does not drive the bill, teams are not penalized for retaining more of it at higher fidelity. An agent needs that depth of evidence to investigate accurately, according to groundcover. The capabilities are available now.

Groundcover is a venture capital-backed startup that has raised $160 million to date, including a $100 million Series C round led by One Peak in July. Its first acquisition came last month, when it bought Kubernetes resource optimization startup Wand Cloud Ltd.

Image: groundcover