Open-source PagerDuty MCP alternatives
PagerDuty's MCP server is itself open source, so you can read what its 64 tools touch before granting access, and it ships read-only by default. The servers here also publish their code, which matters when you want to vet exactly which logs, metrics, or incidents an agent can reach in production.
These are the open-source observability servers people line up against PagerDuty. They lean toward the telemetry side, querying the data behind an alert, where PagerDuty's own strength is routing and on-call. Reading the repo before you wire one in is straightforward for all of them.
The 8 best open-source alternatives
Grafana Labs' open-source server queries dashboards, Prometheus, Loki, incidents, alerts, and OnCall. You can audit the code and still get both telemetry and on-call scheduling in one place.
Set up Grafana →Sentry's server is open source and pulls issues, stack traces, and events while running Seer root-cause analysis. Reading the repo confirms exactly which error data an agent can see.
Set up Sentry →Built on open source from the ground up, the Prometheus server runs PromQL instant and range queries, discovers metrics, and inspects scrape targets, with a small tool surface that is quick to vet.
Set up Prometheus →SigNoz is open source and OpenTelemetry-native: its server reaches traces, logs, metrics, dashboards, and alerts. You can inspect the code before pointing an agent at your telemetry stack.
Set up SigNoz →- DatadogOfficial
Datadog's server is open source even though the platform is hosted. It searches logs, queries metrics, pulls APM traces, inspects monitors, and investigates incidents, with a repo you can read end to end.
Set up Datadog → - HoneycombOfficial
Built for high-cardinality debugging, the open-source Honeycomb server queries traces, metrics, and logs, runs BubbleUp, and manages Boards, Triggers, and SLOs. Auditing it before granting query access is straightforward.
Set up Honeycomb → - New RelicOfficial
New Relic's server is open source and runs NRQL, searches entities, analyzes golden metrics, and triages incidents. You can read how it queries your account before connecting it.
Set up New Relic → Broader than observability, the AWS Labs server runs any AWS CLI command with validation and a read-only mode, and its source is open. Reach for it when an incident touches infrastructure beyond your monitoring tool.
Set up AWS (AWS Labs) →
How to choose
All of these publish their code, so the audit story matches PagerDuty's. For querying telemetry, Grafana, SigNoz, Datadog, and New Relic cover the broadest ground, Prometheus is the focused metrics option, and Sentry and Honeycomb dig into errors and high-cardinality traces. AWS is the outlier, useful when an incident reaches into cloud infrastructure. Read the repo before you grant production access to any of them.
FAQ
- Is the PagerDuty MCP server open source?
- Yes. PagerDuty publishes the server's source, and it runs read-only by default across its 64 tools, so you can audit what it can reach before connecting it. Every alternative on this page also ships its code publicly.
- Why pick an open-source observability MCP server?
- You can read exactly which logs, metrics, traces, or incidents the server is able to touch, pin or patch the version you run, and keep credentials on infrastructure you control. The trade-off is that you take on operating it rather than leaning on the vendor's hosted endpoint.