groundcover, a bring-your-own-cloud (BYOC), eBPF, and OpenTelemetry (OTel)-native observability platform, has announced a multi-surface product launch extending its observability intelligence directly into Agent Mode, external AI coding agents via the Model Context Protocol (MCP), and Slack. The integration bridges the long-standing context gap between writing code and understanding its real-world behavior in production.
groundcover launched agent-to-agent observability, integrating production context into Agent Mode, external IDEs, and Slack.
External coding agents (such as Claude, Cursor, Kiro, and Codex) can now query groundcover’s live telemetry directly via Model Context Protocol (MCP) tools.
The platform supports asynchronous delegated investigations, allowing external agents to hand off deep debugging tasks to the groundcover Agent while continuing their own workflows.
A new Slack integration (groundcover) lets the observability agent join active incident threads, utilize chat history as context, and link with connected tracking systems like Linear and GitHub.
Operating entirely within a customer’s environment via a BYOC and eBPF architecture, the system ensures sensitive production telemetry and prompts never leave local infrastructure.
While modern AI coding assistants excel at reading codebases, writing logic, and reviewing pull requests, they historically lacked visibility into what happened after code reached production. Engineers previously had to manually switch between development environments, ticketing tools, and observability dashboards to correlate code changes with live system telemetry.
groundcover resolves this bottleneck by enabling external AI agents to access production data through two distinct patterns:
Direct Access: When an agent knows what evidence it needs, it can query groundcover’s MCP tools directly for logs, metrics, traces, Kubernetes events, live entities, and active monitors.
Delegated Investigation: When an agent encounters a broader issue without a known root cause, it can hand the task to the groundcover Agent via asynchronous Agent Tasks. groundcover returns a job ID and runs the investigation in the background—performing root cause analysis or generating draft dashboards—while the primary coding agent keeps working.
"Coding agents can tell you what should happen, but they can't tell you what already did happen in production unless they have access to the evidence," said Shahar Azulay, CEO and co-founder of groundcover. "This launch means an agent doesn't have to become an observability expert to get that evidence. It can query groundcover directly or delegate the investigation and keep working while groundcover does the digging."
The new intelligence layer powers three primary entry points for engineering teams:
Agent Mode: groundcover’s native environment for open-ended production investigations, metric tracking, and dynamic dashboard generation.
IDEs and External Agents: Direct integration with tools like Cursor, Kiro, Claude, and Codex via MCP, giving developers runtime observability context inside their command-line and editor interfaces.
Slack Workspace: The groundcover agent can join live incident channels, consume conversation threads as context, and pull in integrated systems like Linear and GitHub to accelerate triage.
About groundcover
groundcover is the full-stack observability platform for engineers & agents, allowing teams to monitor everything they run in their cloud, without compromising on data, control, or cost. It’s deployed inside the customer’s cloud premises and provides complete visibility into applications, infrastructure, and AI workloads without operational overhead. The platform offers unlimited data coverage at a fraction of the cost of legacy observability tools. Founded in 2021, the company is backed by leading global investors, including One Peak, Morgan Stanley Expansion Capital, Zeev Ventures, Angular Ventures, Heavybit and Jibe.
