Superlog MCP
The Superlog MCP server: an agent searches incidents, queries logs, traces and metrics and manages alerts and dashboards.
- 1,480 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- Reads project logs, traces and metrics and, with write rights, changes alerts and dashboards.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/superlog-mcpDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Reads project logs, traces and metrics and, with write rights, changes alerts and dashboards.
This is third-party code. Review the repository files before installing.
What it does
Superlog is an open-source telemetry system on OpenTelemetry. It ingests traces, logs and metrics, groups noise into incidents and watches the application. The MCP server is available at `https://api.superlog.sh/mcp` over Streamable HTTP, sign-in goes through OAuth and rights split into read (`mcp:read`) and write (`mcp:write`). The server code has tools for reading incidents (`get_incident`, `search_incidents`) and alerts (`list_alerts`, `get_alert`), plus queries to telemetry in ClickHouse. Write tools create and change alerts (`create_alert`, `update_alert`) and home dashboard items: widgets, links and layout. Separate settings store project context, error filters and connections to your own MCP servers.
Who it is for. Teams that watch a product through OpenTelemetry and want to work through incidents and change alerts without leaving the editor.
Good fit when
- After a deploy the error share grows and the agent looks for related incidents per service.
- A metric alert with a clear threshold is needed and the agent creates it with the right conditions.
- The on-call engineer asks for a summary of an open incident with logs and traces beside it.
- A team builds a home dashboard from widgets for key services.
Not a fit when
- Telemetry goes to another system and Superlog has no data.
- The agent must not change alerts without approval from the on-call team.
- You need only OpenTelemetry reference with no access to project data.
Example request
Find open incidents for the `checkout` service over the last day and show the logs next to the freshest one.MCP
- Transport
- http
- Authentication
- OAuth
Limitations
It needs a project in Superlog Cloud or your own instance with data. The README describes installing the system itself in detail and says little about the MCP connection: the tool list is taken from the code and the registry, so check it against the site documentation. Write access opens changes to alerts and dashboards, so start with `mcp:read` only. The cloud version has a free tier and paid plans.
How to disable. Revoke the authorization in Superlog settings and remove the MCP connection from the client. Created alerts and widgets are deleted in Superlog itself.
FAQ
How do I connect Superlog to an agent?
Add a remote Streamable HTTP server at `https://api.superlog.sh/mcp` and complete OAuth. To install the system itself in a project the README advises `npx skills add superloglabs/skills --all`.
Can I limit the agent to read-only?
Yes. OAuth scopes split into `mcp:read` and `mcp:write`, and a read-only request closes the tools that create and change alerts.
How do I run Superlog myself?
You need Node.js 20, pnpm 9 and Docker. Run `pnpm install`, `docker compose up -d`, the migrations `pnpm --filter @superlog/db db:migrate` and `pnpm dev`; the web UI opens on port 5173.
What data does the server read?
Incidents, alerts and project telemetry: traces, logs and metrics stored in ClickHouse. Access is limited to the project chosen at authorization.
What data does Superlog ingest?
OpenTelemetry traces, logs and metrics. In the local stack the OTLP intake listens on port 4101, the web UI runs on 5173 and the API on 4100.
Discussion
No questions or reviews yet.
Related
MCP serversThe built-in OneUptime MCP: an agent works incidents, reads logs and traces, acknowledges alerts and sets up monitors.
Datadog MCP Server (winor30)
Datadog MCP Server
Independent Datadog MCP server: an agent reads incidents, monitors, logs, metrics, traces and RUM and schedules downtimes.
Datadog's official managed MCP server: logs, metrics, traces and incidents right in your agent
An agent searches AWS CloudWatch logs, summarizes activity and correlates events across services.
Grafana MCP Server
Grafana MCP server
Grafana's official MCP server: dashboards, Prometheus and Loki queries, alerts and incidents straight from the agent
Observability MCP server: production logs, metrics, traces, exceptions, alerts and deployments readable by the agent.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Superlog MCP (this card) | yes | OAuth | 1,480 | |
| Grafana MCP Server | yes | API key | 3,663 | works with dashboards, Prometheus and Loki in Grafana; Superlog MCP reads incidents and telemetry of its own OpenTelemetry and ClickHouse system |
| Sentry MCP | no | API key | 961 | works through errors and traces in Sentry; Superlog MCP groups signals into incidents and changes alerts |
README in short
The repository holds a web app, an API, an OTLP intake proxy and background workers. The free edition runs on Postgres and ClickHouse, and the built-in agent records a local incident summary.