Yes. The MCP server is free and open source; you only need a Sentry account. Usage of underlying Sentry features (and any embedded LLM provider you wire up for the AI tools) is governed by your own plan and API keys.

Does it support remote/OAuth?

Yes. The canonical deployment is remote at https://mcp.sentry.dev/mcp with OAuth, and it works against both Sentry SaaS and self-hosted instances. A local stdio mode using a user access token is also available.

Sentry MCP server

OfficialSentry712Config last verified Jun 1, 2026

Sentry's official MCP server: pull issues, stack traces, and events, and run Seer root-cause analysis from your editor.

The Sentry MCP server is Sentry's official integration that connects a coding agent to your error tracking and performance data. Instead of switching to the Sentry dashboard, copying a stack trace, and pasting it back, the agent can list organizations, projects, and teams, search and read grouped issues, drill into individual events, inspect tag distributions, and review releases, replays, and profiles. For debugging, it can invoke Seer, Sentry's AI agent, to produce a detailed root-cause analysis with specific code fixes against a production error.

The canonical deployment is the hosted remote endpoint at https://mcp.sentry.dev/mcp, which uses OAuth so there are no long-lived API tokens on disk and works against both Sentry SaaS and self-hosted instances. A local stdio mode is also published as the @sentry/mcp-server npm package, authenticated with a user access token via SENTRY_ACCESS_TOKEN and an optional SENTRY_HOST for self-hosted deployments. The AI-powered search and Seer tools additionally require an embedded LLM provider; all other tools work without one. It exposes a broad, well-named tool surface covering the full issue-to-fix loop.

Quick install

Copy-paste configs are provided for all 8 supported clients. Pick your client below.

Add to ~/.claude.json

~/.claude.json

json

{
  "mcpServers": {
    "sentry": {
      "type": "http",
      "url": "https://mcp.sentry.dev/mcp"
    }
  }
}

Or via CLI

bash

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

Heads up

First tool call opens a browser to authorize.

Step-by-step guides

Available tools

Tool	Description
whoami	Identifies the authenticated user in Sentry.
find_organizations	Finds organizations that the user has access to in Sentry.
find_teams	Finds teams in an organization in Sentry.
find_projects	Finds projects in Sentry.
find_releases	Finds releases in Sentry.
find_dsns	Lists all Sentry DSNs for a specific project.
create_team	Creates a new team in Sentry.
create_project	Creates a new project in Sentry (includes a DSN automatically).
update_project	Updates project settings such as name, slug, platform, and team assignment.
create_dsn	Creates an additional DSN for an existing project.
search_issues	Searches for grouped issues/problems in Sentry, returning a list of issues.
search_events	Searches Sentry events and replays; use for event counts and statistics.
search_issue_events	Searches and filters events within a specific issue.
get_issue_tag_values	Gets the tag value distribution for a specific Sentry issue.
update_issue	Updates a Sentry issue's status or assignment.
get_event_attachment	Downloads attachments from a Sentry event.
get_replay_details	Gets high-level information about a specific Sentry replay by URL or replay ID.
get_profile_details	Inspects a specific Sentry profile in detail.
get_sentry_resource	Fetches a Sentry resource by URL, or by resource type plus resource ID.
analyze_issue_with_seer	Uses Seer to analyze a production error and return root-cause analysis with code fixes.
use_sentry	Natural-language interface to Sentry via an embedded AI agent.
search_docs	Searches Sentry documentation for SDK setup, instrumentation, and configuration guidance.
get_doc	Fetches the full markdown content of a Sentry documentation page.
get_latest_base_snapshot	Gets the latest UI screenshots/images for an app from the preprod snapshot system.
get_snapshot	Gets a preprod snapshot comparison summary including metadata, counts, and changed sections.
get_snapshot_image	Gets metadata and image content for one image in a preprod snapshot.

Required configuration

SENTRY_ACCESS_TOKENOptional
User auth token for the local stdio server. Not needed when using OAuth on the remote endpoint.
SENTRY_HOSTOptional
Host of a self-hosted Sentry instance. Optional; defaults to Sentry SaaS.
OPENAI_API_KEYOptional
API key for the embedded LLM provider required by the AI-powered search and Seer tools (stdio mode).

What you can do with it

Debug a production error without leaving the editor

The agent searches for the relevant issue, reads the stack trace and event details, and runs Seer to get a root-cause analysis with a concrete code fix to apply.

Triage and route incoming issues

Search recent issues across projects, inspect tag distributions to spot affected releases or environments, and update status or assignment to route work to the right owner.

FAQ

Is it free?: Yes. The MCP server is free and open source; you only need a Sentry account. Usage of underlying Sentry features (and any embedded LLM provider you wire up for the AI tools) is governed by your own plan and API keys.
Does it support remote/OAuth?: Yes. The canonical deployment is remote at https://mcp.sentry.dev/mcp with OAuth, and it works against both Sentry SaaS and self-hosted instances. A local stdio mode using a user access token is also available.

← Browse all monitoring-observability servers