About errorwatch
errorwatch provides command-line and MCP access to a core subset of the Sentry API. It is for agents and operators who need to inspect organizations, releases, and issues or update issue status and comments.
Key features
- 01Identify the authenticated token
- 02List organizations, projects, and releases
- 03Search issues with Sentry query syntax
- 04Read issue details and latest events
- 05Update issue status and comments
Use cases
- 01Find unresolved issues in a Sentry project
- 02Review an issue's latest event and comments
- 03Change issue status or add a comment
MCP support
The MCP server exposes tools for Sentry identity, organization and project data, releases, issue search and details, updates, and comments. Calls require `SENTRY_AUTH_TOKEN`.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
11Saved tool list · updated 1w ago
This tool takes no parameters.
Details
From the project READMEerrorwatch
errorwatch keeps an eye on a Sentry account from one small bin. An agent can confirm who a token belongs to, walk organizations, projects, and releases, search the issue stream with Sentry's own query syntax, open an issue and its latest event, change the issue's status, and read or leave comments on it. It runs from the command line or as an MCP server for agents.
A Sentry auth token is required. Put it in the SENTRY_AUTH_TOKEN
environment variable; it is sent to the API as a Bearer token and
never appears in output. When the variable is missing, every call
fails with an error that names it.
This is a core subset of the Sentry API, not the whole surface: the hosted documentation search tools, the AI autofix (Seer) tools, profiling / traces / spans search, DSN and project creation and other admin operations, and organization statistics beyond issues are out of scope.
Install
npx @reneturcios/errorwatch list-organizations
Or run the container image:
docker run -i -e SENTRY_AUTH_TOKEN ghcr.io/cloudcorprecords/errorwatch
The container defaults to the MCP server (see below). Pass CLI arguments to use it as a CLI instead.
CLI
export SENTRY_AUTH_TOKEN="your auth token"
errorwatch whoami
errorwatch list-organizations
errorwatch list-projects --org acme
errorwatch list-releases --org acme --project web
errorwatch search-issues --org acme --query "is:unresolved TypeError"
errorwatch get-issue --issue-id 123456
errorwatch --version
Exit codes: 0 on success, 1 when an API call fails (including a
missing SENTRY_AUTH_TOKEN), 2 for usage errors.
MCP server
One bin serves both faces. The mcp subcommand starts a stdio MCP
server:
npx -y @reneturcios/errorwatch mcp
MCP client configuration:
{
"mcpServers": {
"errorwatch": {
"command": "npx",
"args": ["-y", "@reneturcios/errorwatch", "mcp"],
"env": { "SENTRY_AUTH_TOKEN": "your auth token" }
}
}
}
Tools
| Tool | Input | What it returns |
|---|---|---|
whoami | (none) | The authenticated user or token identity (GET /auth/). |
list_organizations | (none) | Organizations the token can access. |
list_projects | organization_slug (required) | Projects in the organization. |
get_project | organization_slug, project_slug (required) | One project's details. |
list_releases | organization_slug, project_slug (required) | Releases for the project. |
search_issues | organization_slug (required), project_slug, query, limit (optional) | Issues matching the Sentry search query; uses the project issues endpoint when a project is given. |
get_issue | issue_id (required) | One issue's details. |
get_issue_event | issue_id (required), event_id (optional, defaults to latest) | One event belonging to the issue. |
update_issue | issue_id (required), status, resolved_in_next_release, assigned_to, has_seen, is_bookmarked, is_subscribed (optional) | The updated issue; only the fields given are sent. |
list_issue_comments | issue_id (required) | Comments (notes) on the issue. |
create_issue_comment | issue_id, text (required) | The created comment. |
Failed API calls are reported as tool errors carrying the HTTP status
and, when Sentry provides one, its detail message.
Development
npm install
npm run build # tsc -> dist/
npm test # node test runner over test/*.test.ts, no network (fake fetch fixtures)
npm run dev # run the CLI from source via tsx
License
MIT — see LICENSE.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.