Datadog, Inc. launches MCP Server for AI agent access

Datadog announced that its Model Context Protocol (MCP) Server was made generally available to give developers embedding Artificial Intelligence (AI) agents access to live observability data and to support debugging and governed actions within established security and governance controls.

The company said engineering teams were being tasked with operationalizing AI agents and required secure, governed access to production data, reduced integration overhead and compatibility with compliance requirements as agentic systems entered routine use.

Datadog described MCP Server as a purpose-built interface for agentic systems that extended its unified Observability Platform (OP) into AI workflows, feeding live logs, metrics and traces into AI coding agents such as Claude Code, Cursor, Codex, GitHub Copilot, Cognition and Visual Studio Code, and exposing proactive detection and remediation signals to custom agents.

Datadog said MCP Server reduced the risk of breaking changes by providing a dynamic, purpose-built protocol for agent communication and that the server supported debugging, automated remediation and governed access to real-time observability and intelligence.

“By combining telemetry from Datadog’s unified observability platform into teams’ AI workflows, we are enabling the next stage of AI-native development—moving from simply AI copilots to AI operating on live production systems,” said Yanbing Li, Chief Product Officer at Datadog.

The release included forward-looking statements about plans, intentions, expectations, strategies and prospects and cautioned that actual results may differ materially and referenced risks detailed in the company’s Securities and Exchange Commission filings.