Open-source Sentry MCP alternatives
Sentry's server is open source, so you can read how it queries issues, fetches stack traces, and triggers Seer analysis before pointing an agent at your production errors. The alternatives here all publish their code too, which is the reason to prefer them when an agent will reach into live telemetry and you want to know exactly which calls it can make.
Reading the repo also lets you pin a version and patch around an API change rather than wait on a vendor. Every pick below is auditable in that way, though most still authenticate against a hosted backend once you run them.
The 8 best open-source alternatives
Grafana Labs publishes its server, which queries dashboards, Prometheus, Loki, incidents, alerts, and OnCall. With the source open you can confirm what an agent touches across all those data sources.
Set up Grafana →The Prometheus server is open and narrow: PromQL instant and range queries, metric discovery, and scrape-target inspection. A small auditable surface is easy to trust when the data is your own metrics.
Set up Prometheus →SigNoz's open-source server covers traces, logs, metrics, dashboards, and alerts in an OpenTelemetry-native stack. Reading the code matters here because it can both read and create alerts.
Set up SigNoz →PagerDuty's server is open source and exposes incidents, services, and orchestrations across 64 tools, read-only by default. The default plus the visible source make it safe to wire into on-call workflows.
Set up PagerDuty →- DatadogOfficial
Searching logs, querying metrics, pulling APM traces, and investigating incidents, the open-source Datadog server is wide in scope, so being able to audit which endpoints it hits is worth having.
Set up Datadog → - HoneycombOfficial
Honeycomb publishes a server for trace, metric, and log queries plus BubbleUp, Boards, Triggers, and SLOs. The open code lets you see how queries are constructed before an agent runs them.
Set up Honeycomb → - New RelicOfficial
New Relic's open-source hosted server runs NRQL, searches entities, analyzes golden metrics, and triages incidents. You can read how NRQL is built and constrained before granting access.
Set up New Relic → AWS Labs' open-source server runs any AWS CLI command with validation and a read-only mode. It is the pick when the system you watch is the cloud account itself, with the source there to confirm what is allowed.
Set up AWS (AWS Labs) →
How to choose
All of these publish source, so judge by signal and trust model. SigNoz is closest to Sentry for combined error and telemetry coverage; Prometheus and Grafana are the metrics-first picks; Honeycomb is built for trace debugging; Datadog and New Relic are the broad platforms. PagerDuty handles response, and AWS watches the cloud account. Read each repo, and note the read-only defaults on PagerDuty and AWS before connecting.
FAQ
- Is the Sentry MCP server open source?
- Yes. Sentry publishes its server, so you can audit how it pulls issues, stack traces, and events and how Seer root-cause analysis is invoked. Every alternative on this page is open source as well, which lets you read the code before an agent reaches live telemetry.
- Does open source mean these run entirely on my own infrastructure?
- Not necessarily. Open source lets you read and pin the server code, but most of these still authenticate against a hosted backend, Sentry's, Datadog's, New Relic's, when they run. Prometheus and AWS are the ones that point at infrastructure you already operate yourself.