Skip to content
Back to search

toolkit

Agentic tooling for Sentry

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/getsentry/toolkit with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/getsentry/toolkit'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/getsentry--toolkit/badge.svg)](/listings/getsentry--toolkit)

HTML

<a href="/listings/getsentry--toolkit"><img src="/api/listings/getsentry--toolkit/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

You'll find everything you need to know by visiting the deployed service in production:

<https://mcp.sentry.dev>

If you're looking to contribute, learn how it works, or to run this for self-hosted Sentry, continue below.

Claude Code Plugin

Install as a Claude Code plugin for automatic subagent delegation:

claude plugin marketplace add getsentry/sentry-mcp
claude plugin install sentry-mcp@sentry-mcp

This provides a `sentry-mcp` subagent that Claude automatically delegates to when you ask about Sentry errors, issues, traces, or performance.

For forward-looking tool variants and features:

claude plugin install sentry-mcp@sentry-mcp-experimental

Stdio vs Remote

While this repository is focused on acting as an MCP service, we also support a `stdio` transport. This is still a work in progress, but is the easiest way to adapt run the MCP against a self-hosted Sentry install.

**Note:** The AI-powered search tools (`search_errors`, `search_traces`, `search_logs`, `search_issues`, etc.) require an LLM provider (OpenAI, Azure OpenAI, Anthropic, or OpenRouter). These tools use natural language processing to translate queries into Sentry's query syntax. Without a configured provider, these specific tools will be unavailable, but all other tools will function normally.

To utilize the `stdio` transport, you'll need to create an User Auth Token in Sentry with the necessary scopes. As of writing this is:

org:read
project:read
project:write
team:read
team:write
event:write

Launch the transport:

npx @sentry/mcp-server@latest --access-token=sentry-user-token

Need to connect to a self-hosted deployment? Add <code>--host</code> (hostname only, e.g. <code>--host=sentry.example.com</code>) when you run the command. For isolated internal deployments that only expose plain HTTP, also add <code>--insecure-http</code>.

Seer is not part of self-hosted Sentry, so the `seer` skill is left out of the default skill set whenever `--host` points at a non-`sentry.io` host. If your self-hosted deployment does run Seer, opt back in explicitly:

npx @sentry/mcp-server@latest --access-token=TOKEN --host=sentry.example.com --skills=inspect,seer

You can also disable any other skill to prevent unsupported tools from being exposed:

npx @sentry/mcp-server@latest --access-token=TOKEN --host=sentry.example.com --disable-skills=project-management

For self-hosted instances without TLS:

npx @sentry/mcp-server@latest --access-token=TOKEN --host=sentry.internal:9000 --insecure-http

Remote with an Explicit Sentry Token

Remote clients that support custom HTTP headers can pass an upstream Sentry API token directly to the Cloudflare transport:

{
  "mcpServers": {
    "sentry": {
      "url": "https://mcp.sentry.dev/mcp",
      "headers": {
        "Authorization": "Sentry-Bearer ${SENTRY_ACCESS_TOKEN}"
      }
    }
  }
}

`Sentry-Bearer` is intentionally separate from `Bearer`: `Bearer` is reserved for MCP OAuth access tokens. With `Sentry-Bearer`, the worker does not store, validate, exchange, or refresh the upstream token. It forwards the token through the same Sentry API calls used by OAuth-backed sessions, and the client or upstream provider remains responsible for token lifetime and refresh.

Direct remote auth defaults to all active MCP skills. You can narrow the exposed tools with `?skills=inspect,triage` or `?disable-skills=seer`.

Environment Variables

SENTRY_ACCESS_TOKEN=         # Required: Your Sentry auth token

# LLM Provider Configuration (required for AI-powered search tools)
EMBEDDED_AGENT_PROVIDER=     # Required when multiple provider keys are set: 'openai', 'azure-openai', 'anthropic', or 'openrouter'
OPENAI_API_KEY=              # Required if using OpenAI
ANTHROPIC_API_KEY=           # Required if using Anthropic
OPENROUTER_API_KEY=          # Required if using OpenRouter

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 21 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "$schema": "https://anthropic.com/claude-code/marketplace.schema.json",
  3. "name": "sentry-mcp",
  4. "owner": { "name": "Sentry", "email": "oss@sentry.io" },
  5. "plugins": [
  6. {
  7. "name": "sentry-mcp",
  8. "description": "Sentry error tracking and performance monitoring - search issues, analyze errors, explore traces, and get AI-powered root cause analysis",
  9. "author": { "name": "Sentry" },
  10. "source": "./plugins/sentry-mcp",
  11. "category": "monitoring"
  12. },
  13. {
  14. "name": "sentry-mcp-experimental",
  15. "description": "Sentry MCP with experimental features enabled - includes all stable tools plus bleeding-edge capabilities",
  16. "author": { "name": "Sentry" },
  17. "source": "./plugins/sentry-mcp-experimental",
  18. "category": "monitoring"
  19. }
  20. ]
  21. }
README.md · README · 358 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # Sentry Toolkit
  2. This repository contains the Sentry CLI, the Sentry MCP server, and shared
  3. packages. For the CLI, see the [CLI guide](docs/cli/README.md) and the published
  4. documentation at <https://cli.sentry.dev/>. The [documentation index](docs/README.md)
  5. covers both products and shared contributor guidance.
  6. The independent [Docs MCP Worker](apps/docs-mcp/README.md) provides documentation
  7. search and AI-generated answers backed by `docs.sentry.io`.
  8. ## Sentry MCP
  9. Sentry's MCP service is primarily designed for human-in-the-loop coding agents. Our tool selection and priorities are focused on developer workflows and debugging use cases, rather than providing a general-purpose MCP server for all Sentry functionality.
  10. This remote MCP server acts as middleware to the upstream Sentry API, optimized for coding assistants like Cursor, Claude Code, and similar development tools. It's based on [Cloudflare's work towards remote MCPs](https://blog.cloudflare.com/remote-model-context-protocol-servers-mcp/).
  11. ## Getting Started
  12. You'll find everything you need to know by visiting the deployed service in production:
  13. <https://mcp.sentry.dev>
  14. If you're looking to contribute, learn how it works, or to run this for self-hosted Sentry, continue below.
  15. ### Claude Code Plugin
  16. Install as a Claude Code plugin for automatic subagent delegation:
  17. ```shell
  18. claude plugin marketplace add getsentry/sentry-mcp
  19. claude plugin install sentry-mcp@sentry-mcp
  20. ```
  21. This provides a `sentry-mcp` subagent that Claude automatically delegates to when you ask about Sentry errors, issues, traces, or performance.
  22. For forward-looking tool variants and features:
  23. ```shell
  24. claude plugin install sentry-mcp@sentry-mcp-experimental
  25. ```
  26. ### Stdio vs Remote
  27. While this repository is focused on acting as an MCP service, we also support a `stdio` transport. It is the easiest way to run the MCP server against a self-hosted Sentry install.
  28. **Note:** The AI-powered search tools (`search_errors`, `search_traces`, `search_logs`, `search_issues`, etc.) require an LLM provider (OpenAI, Azure OpenAI, Anthropic, or OpenRouter). These tools use natural language processing to translate queries into Sentry's query syntax. Without a configured provider, these specific tools will be unavailable, but all other tools will function normally.
  29. #### Sentry CLI (recommended)
  30. Run the stdio server with the [Sentry CLI](https://cli.sentry.dev)'s `sentry mcp` command. It uses your CLI login, so the MCP config needs no token:
  31. ```shell
  32. curl https://cli.sentry.dev/install -fsS | bash
  33. sentry auth login
  34. claude mcp add sentry -- sentry mcp
  35. ```
  36. The CLI also installs with `brew install getsentry/tools/sentry`,
  37. `npm install -g sentry`, or `irm https://cli.sentry.dev/install.ps1 | iex` on
  38. Windows. See the [CLI installation guide](https://cli.sentry.dev/getting-started/).
  39. Other clients take the same command:
  40. ```json
  41. {
  42. "mcpServers": {
  43. "sentry": {
  44. "command": "sentry",
  45. "args": ["mcp"]
  46. }
  47. }
  48. }
  49. ```
  50. Pass `<org>/` or `<org>/<project>` to limit every call to one organization or
  51. project (`sentry mcp acme/`, `sentry mcp acme/web`), and `--experimental` for
  52. forward-looking tool variants. Run `sentry mcp --help` for every flag and
  53. environment variable, or see the [`sentry mcp` docs](https://cli.sentry.dev/commands/mcp/).
  54. `SENTRY_AUTH_TOKEN`, when set, wins over the stored login. See
  55. [docs/operations/stdio-auth.md](docs/operations/stdio-auth.md).
  56. To use a token, create a User Auth Token in Sentry with the necessary scopes. As of writing this is:
  57. ```
  58. org:read
  59. project:read
  60. project:write
  61. team:read
  62. team:write
  63. event:write
  64. ```
  65. Need to connect to a self-hosted deployment? Log in to it with
  66. `sentry auth login --url https://sentry.example.com`, then set
  67. <code>SENTRY_HOST</code> (e.g. <code>SENTRY_HOST=sentry.example.com</code>)
  68. in the MCP server's environment, which also scopes
  69. <code>SENTRY_AUTH_TOKEN</code> to that server. For isolated internal
  70. deployments that only expose plain HTTP, use an <code>http://</code>
  71. <code>SENTRY_URL</code> instead. See the
  72. [CLI self-hosted guide](https://cli.sentry.dev/self-hosted/).
  73. Seer is not part of self-hosted Sentry, so the `seer` skill is left out of the
  74. default skill set whenever the host is not `sentry.io`. If your
  75. self-hosted deployment does run Seer, opt back in explicitly:
  76. ```shell
  77. SENTRY_HOST=sentry.example.com sentry mcp --skills=inspect,seer
  78. ```
  79. You can also disable any other skill to prevent unsupported tools from being
  80. exposed:
  81. ```shell
  82. SENTRY_HOST=sentry.example.com sentry mcp --disable-skills=project-management
  83. ```
  84. For self-hosted instances without TLS:
  85. ```shell
  86. SENTRY_URL=http://sentry.internal:9000 sentry mcp
  87. ```
  88. #### Legacy: npx @sentry/mcp-server
  89. [`@sentry/mcp-server`](https://www.npmjs.com/package/@sentry/mcp-server) still
  90. ships the same server, bundled without the rest of the CLI, for clients that
  91. launch it through npx. It shares the CLI's login and takes the same flags and
  92. environment variables:
  93. ```shell
  94. npx @sentry/mcp-server@latest auth login # once, to sign in
  95. npx @sentry/mcp-server@latest # uses that login
  96. SENTRY_AUTH_TOKEN=sentry-user-token npx @sentry/mcp-server@latest
  97. ```
  98. `--access-token` / `SENTRY_ACCESS_TOKEN` from older releases still work but are
  99. deprecated: they act as `SENTRY_AUTH_TOKEN` and print a warning on stderr.
  100. #### Remote with an Explicit Sentry Token
  101. Remote clients that support custom HTTP headers can pass an upstream Sentry API
  102. token directly to the Cloudflare transport:
  103. ```json
  104. {
  105. "mcpServers": {
  106. "sentry": {
  107. "url": "https://mcp.sentry.dev/mcp",
  108. "headers": {
  109. "Authorization": "Sentry-Bearer ${SENTRY_ACCESS_TOKEN}"
  110. }
  111. }
  112. }
  113. }
  114. ```
  115. `Sentry-Bearer` is intentionally separate from `Bearer`: `Bearer` is reserved
  116. for MCP OAuth access tokens. With `Sentry-Bearer`, the worker does not store,
  117. validate, exchange, or refresh the upstream token. It forwards the token through
  118. the same Sentry API calls used by OAuth-backed sessions, and the client or
  119. upstream provider remains responsible for token lifetime and refresh.
  120. Direct remote auth defaults to all active MCP skills. You can narrow the exposed
  121. tools with `?skills=inspect,triage` or `?disable-skills=seer`.
  122. #### Environment Variables
  123. ```shell
  124. SENTRY_AUTH_TOKEN= # Optional after `auth login`; wins over a stored login
  125. # LLM Provider Configuration (required for AI-powered search tools)
  126. EMBEDDED_AGENT_PROVIDER= # Required when multiple provider keys are set: 'openai', 'azure-openai', 'anthropic', or 'openrouter'
  127. OPENAI_API_KEY= # Required if using OpenAI
  128. ANTHROPIC_API_KEY= # Required if using Anthropic
  129. OPENROUTER_API_KEY= # Required if using OpenRouter
  130. OPENROUTER_MODEL= # Optional OpenRouter model, defaults to 'openai/gpt-5.6-luna'
  131. OPENROUTER_REASONING_EFFORT= # Optional OpenRouter reasoning effort, defaults to 'high'
  132. # Optional overrides
  133. SENTRY_HOST= # For self-hosted deployments (drops 'seer' from the default skills)
  134. MCP_SKILLS= # Grant specific skills (comma-separated, e.g. 'inspect,seer')
  135. MCP_DISABLE_SKILLS= # Disable specific skills (comma-separated, e.g. 'project-management')
  136. ```
  137. **Important:** Always set `EMBEDDED_AGENT_PROVIDER` to explicitly specify your LLM provider. Auto-detection based on API keys alone is deprecated and will be removed in a future release. See [docs/operations/embedded-agents.md](docs/operations/embedded-agents.md) for detailed configuration options.
  138. #### Example MCP Configuration
  139. ```json
  140. {
  141. "mcpServers": {
  142. "sentry": {
  143. "command": "sentry",
  144. "args": ["mcp"],
  145. "env": {
  146. "EMBEDDED_AGENT_PROVIDER": "openai",
  147. "OPENAI_API_KEY": "sk-..."
  148. }
  149. }
  150. }
  151. }
  152. ```
  153. If you leave the host variable unset, the CLI automatically targets the Sentry
  154. SaaS service. Only set the override when you operate self-hosted Sentry.
  155. Setting `SENTRY_HOST` to a self-hosted host also drops the `seer` skill from
  156. the default set, since Seer is not available on self-hosted Sentry. For a
  157. self-hosted deployment that does run Seer, opt in with `MCP_SKILLS`:
  158. ```json
  159. {
  160. "mcpServers": {
  161. "sentry": {
  162. "command": "sentry",
  163. "args": ["mcp"],
  164. "env": {
  165. "SENTRY_HOST": "sentry.example.com",
  166. "MCP_SKILLS": "inspect,seer"
  167. }
  168. }
  169. }
  170. }
  171. ```
  172. ### Rate Limits
  173. Tool calls are subject to the Sentry API's rate limits. The MCP server retries
  174. rate-limited requests, and transient Sentry server errors on read requests,
  175. after a short wait. When a limit lasts longer, the tool returns a **Rate
  176. Limited** message that says when to try again. Agents hit the per-second
  177. limits most often by running many Sentry tool calls in parallel, so run them
  178. one at a time if you see this message.
  179. ### MCP Inspector
  180. MCP includes an [Inspector](https://modelcontextprotocol.io/docs/tools/inspector), to easily test the service:
  181. ```shell
  182. pnpm inspector
  183. ```
  184. Enter the MCP server URL (<http://localhost:5173>) and hit connect. This should trigger the authentication flow for you.
  185. Note: If you have issues with your OAuth flow when accessing the inspector on `127.0.0.1`, try using `localhost` instead by visiting `http://localhost:6274`.
  186. ## Local Development
  187. To contribute changes, you'll need to set up your local environment:
  188. 1. **Set up environment and agent skills:**
  189. ```shell
  190. make setup-env # Creates .env files and installs shared agent skills
  191. ```
  192. This also runs `npx @sentry/dotagents install` to install shared skills from [getsentry/skills](https://github.com/getsentry/skills) into `.agents/skills/` (symlinked into `.claude/skills` and `.cursor/skills`). If you need to update skills later, run it directly:
  193. ```shell
  194. npx @sentry/dotagents install
  195. ```
  196. 2. **Create an OAuth App in Sentry** (Settings => API => [Applications](https://sentry.io/settings/account/api/applications/)):
  197. - Homepage URL: `http://localhost:5173`
  198. - Authorized Redirect URIs: `http://localhost:5173/oauth/callback`
  199. - Note your Client ID and generate a Client secret
  200. 3. **Configure your credentials:**
  201. - Edit `.env` in the root directory and add either `OPENAI_API_KEY` or `OPENROUTER_API_KEY`
  202. - Edit `packages/mcp-cloudflare/.env` and add:
  203. - `SENTRY_CLIENT_ID=your_development_sentry_client_id`
  204. - `SENTRY_CLIENT_SECRET=your_development_sentry_client_secret`
  205. - `COOKIE_SECRET=my-super-secret-cookie`
  206. 4. **Start the development server:**
  207. ```shell
  208. pnpm dev
  209. ```
  210. ### Verify
  211. Run the server locally to make it available at `http://localhost:5173`
  212. ```shell
  213. pnpm dev
  214. ```
  215. To test the local server, enter `http://localhost:5173/mcp` into Inspector and hit connect. Once you follow the prompts, you'll be able to "List Tools".
  216. ### Tests
  217. There are three test suites included: unit tests, evaluations, and manual testing.
  218. **Unit tests** can be run using:
  219. ```shell
  220. pnpm test
  221. ```
  222. **Evaluations** require a `.env` file in the project root with some config:
  223. ```shell
  224. # .env (in project root)
  225. OPENAI_API_KEY= # Use OpenAI-backed AI-powered tools
  226. OPENROUTER_API_KEY= # Or use OpenRouter-backed AI-powered tools
  227. ```
  228. Note: The root `.env` file provides defaults for all packages. Individual packages can have their own `.env` files to override these defaults during development.
  229. Once that's done you can run them using:
  230. ```shell
  231. pnpm eval
  232. ```
  233. **Manual testing** (preferred for testing MCP changes):
  234. ```shell
  235. # Test with local dev server (default: http://localhost:5173)
  236. pnpm -w run cli "who am I?"
  237. # Test against production
  238. pnpm -w run cli --mcp-host=https://mcp.sentry.dev "query"
  239. # Test with local stdio mode (requires SENTRY_ACCESS_TOKEN)
  240. pnpm -w run cli --access-token=TOKEN "query"
  241. ```
  242. Note: The CLI defaults to `http://localhost:5173`. Override with `--mcp-host` or set `MCP_URL` environment variable.
  243. **Comprehensive testing playbooks:**
  244. - **Stdio testing:** See `docs/testing/stdio.md` for complete guide on building, running, and testing the stdio implementation (IDEs, MCP Inspector)
  245. - **Remote testing:** See `docs/testing/remote.md` for complete guide on testing the remote server (OAuth, web UI, CLI client)
  246. ## Development Notes
  247. ### Automated Code Review
  248. This repository uses automated code review tools (like Cursor BugBot) to help identify potential issues in pull requests. These tools provide helpful feedback and suggestions, but **we do not recommend making these checks required** as the accuracy is still evolving and can produce false positives.
  249. The automated reviews should be treated as:
  250. - ✅ **Helpful suggestions** to consider during code review
  251. - ✅ **Starting points** for discussion and improvement
  252. - ❌ **Not blocking requirements** for merging PRs
  253. - ❌ **Not replacements** for human code review
  254. When addressing automated feedback, focus on the underlying concerns rather than strictly following every suggestion.
  255. ### Contributor Documentation
  256. Looking to contribute? Start at the [Toolkit documentation index](docs/README.md)
  257. for CLI, MCP, and shared guides. See `AGENTS.md` for repository rules and
  258. `packages/cli/AGENTS.md` for CLI-specific rules.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.