Coralogix MCP

Coralogix's MCP server exposes its observability data and its configuration surfaces to an MCP client, so an agent can both investigate an incident and adjust the pipeline that produced the evidence. Queries run against logs, metrics and traces using Coralogix query syntax, with the usual filters on time range, application and subsystem, which lets an agent narrow from a noisy alert down to the specific spans or log lines behind it. Configuration matters as much as the data here. Agents can work with alert definitions, and with the parsing rules that turn raw log lines into structured fields, plus the settings that decide how incoming telemetry is routed and retained. That makes the server useful for the ongoing noise and volume work observability teams do between incidents, not only for firefighting during one. An agent can be asked why a field stopped populating and trace the answer back to the rule that changed. Access is authenticated with a Coralogix API key scoped to a team, so what an agent can read or modify follows permissions already defined in the account rather than a separate grant. Built for SRE, DevOps and platform engineers who want alert triage, query building and rule debugging happening inside a coding assistant rather than in a separate console tab.