Sentry

# Sentry

Read errors, stack traces and release health from Sentry, and create/resolve issues. Read-only for triage, write-capable for issue management.

- **Publisher of the server:** Sentry
- **Upstream:** https://docs.sentry.io/product/sentry-mcp/
- **Documentation:** https://docs.sentry.io/product/sentry-mcp/
- **Transport:** streamable HTTP (remote)
- **Endpoint:** `https://mcp.sentry.dev/mcp`

## Credentials and configuration

- Authorization: Bearer <Sentry auth token>

These are **placeholders describing what to supply**; the registry stores no credentials. Enter real values in your own AO instance when you create the connection.

## Tools (6)

- `list_issues` — List issues matching a query.
- `get_issue_details` — Read one issue's events and stack traces.
- `create_issue` — Create a Sentry issue.
- `resolve_issue` — Mark an issue resolved.
- `search_events` — Search raw error events.
- `analyze_severity` — Summarise an issue's severity trend over time.

## Risk

Declared risk: **medium**. Tools that can change state: `create_issue`, `resolve_issue`. Write access must be enabled explicitly.

> This entry is a catalogue description. Installing it creates no connection and grants no agent access;
> your instance's own connection settings and approvals decide what an agent can reach.