Superlog

What Superlog MCP server does

The Superlog MCP server provides an agent-facing way to work with telemetry and incident data from Superlog. Its supported scope covers logs, traces, metrics, and incidents. Superlog is an open-core observability workspace built around OpenTelemetry data, so it is intended for teams investigating application and infrastructure behavior rather than for general-purpose data access.

The underlying community edition includes a web application and API, an OTLP ingest proxy, worker processes, database migrations, and telemetry query support backed by Postgres and ClickHouse. It also includes agent runner interfaces and a default community runner that records a local incident summary.

How it works

Telemetry enters Superlog through its OTLP intake path. The system stores and processes the incoming traces, logs, and metrics, then groups noisy signals into incidents. An agent can use the resulting observability context to inspect production behavior and work from incident-level information instead of treating every signal as unrelated.

The repository separates the main parts of the stack into web, API, proxy, and worker applications, with database and telemetry-fingerprinting packages supporting them. This structure indicates that the Superlog MCP server is best suited to workflows where an agent needs an observability workspace behind its queries, rather than a standalone log-file reader.

Setup and configuration

The documented installation path uses Superlog skills with a coding agent:

npx skills add superloglabs/skills --all

The repository’s local development setup requires Node.js 20 or newer, pnpm 9 or newer, and Docker. After installing dependencies, start the local services with Docker Compose, apply the database migrations, and launch the development process:

pnpm install
docker compose up -d
pnpm --filter @superlog/db db:migrate
pnpm dev

The local web application uses port 5173, the API uses port 4100, and OTLP intake uses port 4101.

Tools and capabilities

The Superlog MCP server is relevant when an agent needs to:

• Query application logs.
• Inspect distributed traces.
• Examine collected metrics.
• Review incidents created from related telemetry.
• Use incident summaries produced by the community agent runner.

Superlog supports both local deployment and a hosted cloud edition. The repository described here is the free, fully open-source community edition; the hosted product is mentioned separately and is not required for the local stack.

Limitations and notes

The available documentation does not list individual MCP tool names, request schemas, authentication settings, or supported client applications. It also does not specify whether the MCP interface uses stdio, HTTP, or another transport. Confirm those details in the repository before building a client-specific integration.

The local setup is a multi-service observability system, not just a single executable. Running it requires Docker, Node.js, pnpm, database migrations, and the development processes described above. The Apache License 2.0 applies to the repository.