> ## Documentation Index
> Fetch the complete documentation index at: https://opensre.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Sentry (MCP)

> Connect Sentry's hosted MCP server so OpenSRE can query issues, events, traces, releases, and Seer root-cause analysis during investigations

OpenSRE connects to Sentry's hosted [Model Context Protocol (MCP)](https://mcp.sentry.dev) server, exposing Sentry's products — issues, events, traces, replays, releases, monitors, and Seer AI root-cause analysis — as tools the agent can call while investigating an incident.

This is distinct from the [Sentry issue integration](/sentry), which is a narrow REST client used for issue and event lookup. Use the MCP integration when you want the agent to explore Sentry data directly, including Seer-assisted debugging.

## Tools

| Tool                | What it does                                                                                      |
| ------------------- | ------------------------------------------------------------------------------------------------- |
| `list_sentry_tools` | List the tools the connected Sentry MCP server exposes                                            |
| `call_sentry_tool`  | Call a named Sentry MCP tool (e.g. look up an issue, fetch a trace, run Seer root-cause analysis) |

The agent typically calls `list_sentry_tools` first to discover what is available for your organization, then `call_sentry_tool` with the chosen tool name and arguments.

## Prerequisites

* A Sentry account (hosted SaaS, or self-hosted via the `SENTRY_MCP_HOST` override)
* A Sentry **user auth token** created in your account settings. The required scopes depend on the skills you use:
  * `org:read` — read-only inspection of issues, events, traces, releases, monitors (default)
  * `event:write`, `team:write`, `project:write` — for triage and project-management skills

## Setup

### Option 1: Interactive CLI

```bash theme={null}
opensre integrations setup
```

Select **Sentry (MCP)** when prompted, then paste your user auth token. The setup uses the hosted Streamable HTTP transport; keep the default URL unless you run self-hosted Sentry. To run a local server instead, set `SENTRY_MCP_MODE=stdio` via environment variables (see below).

### Option 2: Environment variables

Add to your `.env`:

```bash theme={null}
SENTRY_MCP_MODE=streamable-http
SENTRY_MCP_URL=https://mcp.sentry.dev/mcp
SENTRY_MCP_AUTH_TOKEN=your_sentry_user_auth_token
SENTRY_MCP_HOST=                       # optional, self-hosted Sentry host
SENTRY_MCP_ORGANIZATION_SLUG=          # optional, scope to one organization
SENTRY_MCP_PROJECT_SLUG=               # optional, scope to one project
SENTRY_MCP_SKILLS=                     # optional, comma-separated skill filter
```

| Variable                       | Default                      | Description                                                                      |
| ------------------------------ | ---------------------------- | -------------------------------------------------------------------------------- |
| `SENTRY_MCP_AUTH_TOKEN`        | —                            | **Required** (hosted). Sentry user auth token                                    |
| `SENTRY_MCP_URL`               | `https://mcp.sentry.dev/mcp` | MCP server URL                                                                   |
| `SENTRY_MCP_MODE`              | `streamable-http`            | Transport: `streamable-http`, `sse`, or `stdio`                                  |
| `SENTRY_MCP_HOST`              | —                            | Self-hosted Sentry hostname (e.g. `sentry.example.com`)                          |
| `SENTRY_MCP_ORGANIZATION_SLUG` | —                            | Scope tools to a specific organization                                           |
| `SENTRY_MCP_PROJECT_SLUG`      | —                            | Scope tools to a specific project                                                |
| `SENTRY_MCP_SKILLS`            | —                            | Comma-separated skill filter (`inspect`, `seer`, `triage`, `project-management`) |
| `SENTRY_MCP_COMMAND`           | —                            | Command to launch a local MCP server (`stdio` mode only)                         |
| `SENTRY_MCP_ARGS`              | —                            | Arguments for the local MCP command (`stdio` mode only)                          |

To run a local Sentry MCP server instead of the hosted endpoint, use `stdio` mode:

```bash theme={null}
SENTRY_MCP_MODE=stdio
SENTRY_MCP_COMMAND=npx
SENTRY_MCP_ARGS=@sentry/mcp-server@latest
SENTRY_MCP_AUTH_TOKEN=your_sentry_user_auth_token
```

<Info>
  The `search_events` and `search_issues` tools translate natural-language queries and require an `OPENAI_API_KEY` in the MCP server's environment. The rest of the MCP server works without it, so you can skip this if you do not need those tools.
</Info>

### Option 3: Persistent store

```json theme={null}
{
  "version": 1,
  "integrations": [
    {
      "id": "sentry-mcp-prod",
      "service": "sentry_mcp",
      "status": "active",
      "credentials": {
        "url": "https://mcp.sentry.dev/mcp",
        "mode": "streamable-http",
        "auth_token": "your_sentry_user_auth_token",
        "organization_slug": "my-org"
      }
    }
  ]
}
```

## Verify

```bash theme={null}
opensre integrations verify sentry_mcp
```

A successful check connects to the MCP server and reports how many tools it discovered. If it fails, the most common cause is a missing or invalid user auth token — confirm the token has at least `org:read` scope and that outbound HTTPS to `mcp.sentry.dev` is allowed.