Skip to content
Menu

MCP server / observability / source-linked

Sentry MCP

Sentry MCP is a hosted, human-in-the-loop debugging server optimized for coding agents. It connects to Sentry through OAuth and can search errors, inspect performance, triage issues, read documentation, and manage projects. Organization/project URL scoping hides irrelevant discovery tools and is recommended for safer context.

Before you choose.

Identify the implementation and the job it supports.

Which implementation is this?
Sentry MCP
Implementation source ↗

Sentry MCP is a hosted, human-in-the-loop debugging server optimized for coding agents. It connects to Sentry through OAuth and can search errors, inspect performance, triage issues, read documentation, and manage projects. Organization/project URL scoping hides irrelevant discovery tools and is recommended for safer context.

Setup and connection.

The documented interface, authentication and installation path.

Connection / transport
Streamable HTTP
Authentication
OAuth for the remote endpoint; a user auth token is documented for the in-progress stdio/self-hosted path.

Use the scoped https://mcp.sentry.dev/mcp/{organization}/{project} endpoint in a supported client and complete OAuth. For self-hosted stdio, install the package and pass a token.

Setup reference ↗

Documented setup example. Review the source and credentials before running; this command was not executed for this article.

claude mcp add --transport http sentry https://mcp.sentry.dev/mcp/ORG/PROJECT

What it actually exposes.

Concrete capabilities with maintainer sources.

Documented capability / 01

Error investigation

Search issues and events and inspect traces or performance signals.

Documentation ↗
Documented capability / 02

Project operations

Read or manage project information through the remote service.

Documentation ↗
Documented capability / 03

Agent handoff

The official repository includes a Sentry-specific Claude subagent plugin for error and trace questions.

Documentation ↗

Evidence and method.

What was checked and what was not tested.

WHO

Editorial record

Compiled by clawbot.ai.

HOW

Source method

Public maintainer documentation and repositories. This is not a hands-on product review.

WHEN

Last checked 2026-10-01

Setup examples are documented examples. Suggested trials and comparison advice are editorial guidance.

Sources support the adjacent factual statements. A source review does not establish uptime, performance, security or current account eligibility. Read the editorial policy.

Try a specific task.

A bounded evaluation you can reproduce in your own environment.

Production error triage

Suggested evaluation, not an observed test result.

  1. Connect the smallest Sentry project scope and authorize OAuth.
  2. Ask for the newest grouped errors, affected releases, and representative traces.
  3. Draft a fix issue and require a human to approve any project write.

What to inspect: A scoped incident brief linking symptoms, release context, and candidate next actions. Record the client, selected scope, date, and approval checkpoint so readers can reproduce the decision. Keep the source links beside each claim when publishing the record. Proposed workflow; not tested by clawbot.ai.

Workflow reference ↗

Permissions and limitations.

Read these constraints before granting access.

Boundary / 01

AI-powered search tools in stdio require an LLM provider; they are unavailable without one.

Boundary reference ↗
Boundary / 02

Self-hosted stdio is described as work in progress and needs explicit token scopes; do not equate it with the hosted OAuth service.

Boundary reference ↗

Decide by workflow.

Make the expected output explicit before you connect.

A scoped incident brief linking symptoms, release context, and candidate next actions. Record the client, selected scope, date, and approval checkpoint so readers can reproduce the decision. Keep the source links beside each claim when publishing the record. Proposed workflow; not tested by clawbot.ai.

Keep the task inputs, tool calls, output and any errors together when you evaluate. The trial above is a proposal, not a recorded result.

Compare the alternatives.

Specific reasons to choose another route, with the tradeoff shown.

Editorial decision guidance inferred from the linked documentation; no quality or speed benchmark.
AlternativeChoose it whenTradeoff
Grafana MCPChoose when dashboards, logs, traces, and datasource queries span an observability stack.Grafana requires managing instance credentials and server deployment for local use.
Documentation ↗
Vercel MCPChoose when the failure is tied specifically to Vercel deployments and runtime logs.Vercel scope is platform-specific; Sentry offers broader error context.
Documentation ↗

Questions before setup.

Answers specific to this implementation.

Is this a general Sentry API server?

No. Sentry says the service prioritizes human-in-the-loop developer debugging workflows.
Documentation ↗

Can I scope a connection?

Yes. Include organization and project slugs in the endpoint; scoped connections hide unnecessary discovery tools.
Documentation ↗