sentry-reference-architecture
Establishes best practices for structuring Sentry projects across large, multi-service organizations.
Install
mkdir -p .claude/skills/sentry-reference-architecture && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/9037" && unzip -o skill.zip -d .claude/skills/sentry-reference-architecture && rm skill.zipInstalls to .claude/skills/sentry-reference-architecture
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Design production-grade Sentry architecture for multi-service organizations.Key capabilities
- →Implement one-project-per-service topology
- →Create centralized Sentry configuration modules
- →Configure team-based alert routing
- →Set up distributed tracing across microservices
- →Manage error middleware for Express and FastAPI
How it works
It establishes a centralized configuration module to enforce organization-wide defaults and project-specific topology for multi-service environments.
Inputs & outputs
When to use sentry-reference-architecture
- →Plan Sentry project rollout
- →Set up distributed tracing for microservices
- →Configure team-based alert routing
- →Implement centralized Sentry configuration
About this skill
Sentry Reference Architecture
Overview
Enterprise Sentry architecture patterns for multi-service organizations. Covers centralized configuration, project topology, team-based alert routing, distributed tracing, error middleware, source map management, and a production-ready SentryService wrapper.
Prerequisites
- Sentry organization at sentry.io (Business plan+ for team features)
@sentry/nodev8+ installed (npm install @sentry/node @sentry/profiling-node)- Service inventory and team ownership documented
- Node.js 18+ (ESM and native fetch instrumentation)
Instructions
Step 1 — Project Structure Strategy
Pattern A: One Project Per Service (3+ services, recommended)
Organization: acme-corp
├── Team: platform-eng
│ ├── Project: api-gateway (Node/Express)
│ ├── Project: auth-service (Node/Fastify)
│ └── Project: user-service (Node/Express)
├── Team: payments
│ ├── Project: payment-api (Node/Express)
│ └── Project: billing-worker (Node worker)
└── Team: frontend
├── Project: web-app (React/Next.js)
└── Project: mobile-app (React Native)
Benefits: independent quotas, team-scoped alerts, per-service rate limits, isolated release tracking.
Pattern B: Shared Project (< 3 services, single team) — one project with Environment tags (production/staging/dev). Simpler setup; outgrow when alert noise exceeds one team.
Step 2 — Centralized Config Module
Create lib/sentry.ts imported by every service to enforce org-wide defaults:
// lib/sentry.ts
import * as Sentry from '@sentry/node';
import { nodeProfilingIntegration } from '@sentry/profiling-node';
export interface SentryServiceConfig {
serviceName: string;
dsn: string;
environment?: string;
version?: string;
tracesSampleRate?: number;
ignoredTransactions?: string[];
}
export function initSentry(config: SentryServiceConfig): void {
const env = config.environment || process.env.NODE_ENV || 'development';
Sentry.init({
dsn: config.dsn,
environment: env,
release: `${config.serviceName}@${config.version || 'unknown'}`,
serverName: config.serviceName,
tracesSampleRate: config.tracesSampleRate ?? (env === 'production' ? 0.1 : 1.0),
sendDefaultPii: false,
maxBreadcrumbs: 50,
integrations: [nodeProfilingIntegration()],
ignoreErrors: [
'ResizeObserver loop completed with undelivered notifications',
/Loading chunk \d+ failed/,
'AbortError',
],
tracesSampler: ({ name, parentSampled }) => {
if (parentSampled !== undefined) return parentSampled;
const ignored = config.ignoredTransactions || [
'GET /health', 'GET /healthz', 'GET /ready', 'GET /metrics',
];
if (ignored.some(p => name.includes(p))) return 0;
return config.tracesSampleRate ?? (env === 'production' ? 0.1 : 1.0);
},
beforeSend(event) {
if (event.request?.headers) {
delete event.request.headers['authorization'];
delete event.request.headers['cookie'];
delete event.request.headers['x-api-key'];
}
return event;
},
initialScope: {
tags: { service: config.serviceName, team: process.env.TEAM_NAME || 'unassigned' },
},
});
}
export { Sentry };
Bootstrap in each service:
import { initSentry } from '@acme/sentry-config';
initSentry({ serviceName: 'api-gateway', dsn: process.env.SENTRY_DSN! });
// Must run BEFORE other imports that need instrumentation
Step 3 — Error Handling Middleware
Express:
// lib/sentry-middleware.ts
import * as Sentry from '@sentry/node';
import type { Request, Response, NextFunction, ErrorRequestHandler } from 'express';
export function sentryRequestHandler() {
return (req: Request, _res: Response, next: NextFunction): void => {
Sentry.setTag('http.route', req.route?.path || req.path);
if (req.user) Sentry.setUser({ id: req.user.id });
next();
};
}
export const sentryErrorHandler: ErrorRequestHandler = (err, req, res, _next) => {
const status = (err as any).statusCode || 500;
Sentry.withScope((scope) => {
scope.setLevel(status >= 500 ? 'error' : 'warning');
scope.setTag('http.status_code', String(status));
scope.setContext('request', { method: req.method, url: req.originalUrl });
status >= 500 ? Sentry.captureException(err) : Sentry.captureMessage(err.message, 'warning');
});
res.status(status).json({ error: status >= 500 ? 'Internal server error' : err.message });
};
// Wire: app.use(sentryRequestHandler()) → routes → app.use(sentryErrorHandler)
FastAPI (Python):
import sentry_sdk, os
from sentry_sdk.integrations.fastapi import FastApiIntegration
def init_sentry(service_name: str) -> None:
sentry_sdk.init(
dsn=os.environ["SENTRY_DSN"],
environment=os.getenv("ENV", "development"),
release=f"{service_name}@{os.getenv('SERVICE_VERSION', 'unknown')}",
traces_sample_rate=0.1,
send_default_pii=False,
integrations=[FastApiIntegration(transaction_style="endpoint")],
before_send=lambda event, hint: _scrub(event),
)
def _scrub(event):
headers = event.get("request", {}).get("headers", {})
for k in ["authorization", "cookie", "x-api-key"]:
headers.pop(k, None)
return event
Step 4 — Distributed Tracing
HTTP (automatic): SDK v8 auto-propagates sentry-trace + baggage headers on fetch/http. All services in the same org link automatically.
Message queues (manual propagation):
// lib/sentry-queue.ts
import * as Sentry from '@sentry/node';
export function publishWithTrace<T>(queue: string, payload: T, publish: Function) {
return Sentry.startSpan({ name: `queue.publish.${queue}`, op: 'queue.publish' }, async () => {
const headers: Record<string, string> = {};
const span = Sentry.getActiveSpan();
if (span) {
headers['sentry-trace'] = Sentry.spanToTraceHeader(span);
headers['baggage'] = Sentry.spanToBaggageHeader(span) || '';
}
await publish(queue, { payload, headers });
});
}
export function consumeWithTrace<T>(queue: string, msg: { payload: T; headers: Record<string, string> }, handler: (p: T) => Promise<void>) {
return Sentry.continueTrace(
{ sentryTrace: msg.headers['sentry-trace'], baggage: msg.headers['baggage'] },
() => Sentry.startSpan({ name: `queue.process.${queue}`, op: 'queue.process' }, () => handler(msg.payload))
);
}
Step 5 — Team-Based Alert Routing
Configure in Project Settings > Ownership Rules:
# .sentry/ownership-rules
path:src/payments/** #payments-team
path:src/auth/** #platform-team
url:*/api/v1/payments/* #payments-team
tags.service:payment-api #payments-team
* #platform-team
Alert tiers:
- P0 Critical: error rate > 50/min OR crash-free < 95% — PagerDuty on-call, 15 min SLA
- P1 Warning: new production issue or regression — Slack #alerts-prod, same-day SLA
- P2 Performance: P95 > 2s or Apdex < 0.7 — Slack #alerts-perf, next sprint
- P3 Info: new staging issue — Slack #alerts-staging, backlog triage
Step 6 — Source Map Uploads (Monorepo)
#!/usr/bin/env bash
# scripts/upload-sourcemaps.sh — run in CI after build
set -euo pipefail
SERVICE="${1:?Usage: upload-sourcemaps.sh <service>}"
RELEASE="${SERVICE}@$(git rev-parse --short HEAD)"
npx @sentry/cli releases new "$RELEASE" --org "$SENTRY_ORG" --project "$SERVICE"
npx @sentry/cli sourcemaps upload --org "$SENTRY_ORG" --project "$SERVICE" \
--release "$RELEASE" --url-prefix "~/" --validate "./services/${SERVICE}/dist/"
npx @sentry/cli releases set-commits "$RELEASE" --org "$SENTRY_ORG" --auto
npx @sentry/cli releases finalize "$RELEASE" --org "$SENTRY_ORG"
Webpack plugin alternative (auto-uploads on build):
import { sentryWebpackPlugin } from '@sentry/webpack-plugin';
export default {
devtool: 'source-map',
plugins: [sentryWebpackPlugin({
org: process.env.SENTRY_ORG,
project: 'web-app',
authToken: process.env.SENTRY_AUTH_TOKEN,
sourcemaps: { filesToDeleteAfterUpload: ['./dist/**/*.map'] },
})],
};
Step 7 — Custom Integrations
Wrap internal SDK calls with spans for tracing visibility:
import * as Sentry from '@sentry/node';
export function withSpan<T>(op: string, desc: string, fn: () => Promise<T>, attrs?: Record<string, string | number>): Promise<T> {
return Sentry.startSpan({ name: desc, op, attributes: attrs }, async (span) => {
try { const r = await fn(); span.setStatus({ code: 1, message: 'ok' }); return r; }
catch (e) { span.setStatus({ code: 2, message: String(e) }); throw e; }
});
}
// Usage: await withSpan('payment.charge', 'charge $50', () => gateway.charge(5000, 'usd', custId));
Step 8 — SentryService Wrapper
Production wrapper with singleton, metrics, and graceful shutdown:
// lib/sentry-service.ts
import * as Sentry from '@sentry/node';
export class SentryService {
private static instance: SentryService | null = null;
private initialized = false;
private constructor(private config: { serviceName: string; dsn: string; version?: string }) {}
static getInstance(config?: { serviceName: string; dsn: string; version?: string }): SentryService {
if (!SentryService.instance) {
if (!config) throw new Error('Config required on first call');
SentryService.instance = new SentryService(config);
}
return SentryService.instance;
}
init(): void {
if (this.initialized) return;
const { initSentry } = require('./sentry');
initSentry(this.config);
this.initialized = true;
}
captureError(error: Error, ctx?: { tags?: Record<string, string>; extra?: Record<string, unknown>; level?: Sentry.SeverityLevel }): string {
return Sentry.withScope((scope) => {
if (ctx?.tags) Object.entries(ctx.tag
---
*Content truncated.*
When not to use it
- →When managing fewer than 3 services with a single team
Prerequisites
Limitations
- →Alert noise exceeds one team in shared projects
- →Ownership rules mismatch can route alerts to wrong teams
How it compares
This approach uses a shared configuration module to enforce consistency across services rather than manual, per-service initialization.
Compared to similar skills
sentry-reference-architecture side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| sentry-reference-architecture (this skill) | 0 | 26d | Review | Advanced |
| software-architecture | 333 | 6mo | No flags | Intermediate |
| architect-review | 109 | 4mo | No flags | Advanced |
| mcp-builder | 136 | 3mo | Review | Advanced |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by jeremylongshore
View all by jeremylongshore →You might also like
software-architecture
davila7
Guide for quality focused software architecture. This skill should be used when users want to write code, design architecture, analyze code, in any case that relates to software development.
architect-review
sickn33
Master software architect specializing in modern architecture patterns, clean architecture, microservices, event-driven systems, and DDD. Reviews system designs and code changes for architectural integrity, scalability, and maintainability. Use PROACTIVELY for architectural decisions.
mcp-builder
anthropics
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
solid-principles
SmidigStorm
Enforce SOLID principles (Single Responsibility, Open/Closed, Liskov Substitution, Interface Segregation, Dependency Inversion) in object-oriented design. Use when writing or reviewing classes and modules.
codex
Lucklyric
Invoke Codex CLI for complex coding tasks requiring high reasoning capabilities. This skill should be invoked when users explicitly mention "Codex", request complex implementation challenges, advanced reasoning, or need high-reasoning model assistance. Automatically triggers on codex-related requests and supports session continuation for iterative development.
architecture-patterns
wshobson
Implement proven backend architecture patterns including Clean Architecture, Hexagonal Architecture, and Domain-Driven Design. Use when architecting complex backend systems or refactoring existing applications for better maintainability.