agentleFS
mejbaurbaharmejbaurbahar/software-tester-skills/skills/production-monitoring-synthetic-testing/SKILL.md

production-monitoring-synthetic-testing

Skill1 starsChanged 57 days agoMITView on GitHub

Use when testing in production safely (shift-right) — synthetic monitoring, canary checks, uptime probes, Playwright/Checkly/Datadog synthetics, real-user monitoring, SLO/SLI validation, alert testing, dark launches, traffic shadowing and observability-driven QA.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read-mostly journeys; writes use sandbox payment mode/void afterwards.

    line 41
  2. Distinctive:

    No load tests against prod unless planned/throttled.

    line 42
  3. Distinctive:

    Every incident → (1) regression test, (2) new/tuned monitor, (3) runbook fix.

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Production Monitoring & Synthetic Testing (Shift-Right)

Pre-prod can't reproduce real data, traffic, third parties or geography. Continuously prove production works from the user's point of view, safely.

Synthetic monitor types

TypeFrequencyExample
Uptime/HTTP30–60 s, multi-regionGET /healthz 200 + body assertion, TLS expiry > 14 d
API transaction1–5 minlogin → create → read → delete with test tenant
Browser journey5–15 minPlaywright: search → add to cart → checkout (to payment sandbox stub)
DNS/CDN/cert5 minrecords, propagation, chain validity
Third-party dependency5 minpayment/email/auth provider reachability
Cron/heartbeatper jobdead-man's-switch (Healthchecks.io, Cronitor)

Tools: Checkly (Playwright-native), Datadog/New Relic/Grafana Synthetic, Pingdom, UptimeRobot, Better Stack, k6 Cloud browser, Sentry Crons, self-hosted Playwright in CI cron.

// Checkly/Playwright synthetic — runs every 10 min from 3 regions
import { test, expect } from '@playwright/test';
test('critical: user can search and reach checkout', async ({ page }) => {
  await page.goto('https://shop.example.com');
  await page.getByRole('searchbox').fill('headphones'); await page.keyboard.press('Enter');
  await page.getByRole('link', { name: /headphones/i }).first().click();
  await page.getByRole('button', { name: 'Add to cart' }).click();
  await expect(page.getByTestId('cart-count')).toHaveText('1');
});

Safe testing in production

  • Dedicated test tenant/accounts, flagged so they're excluded from analytics, billing, emails, inventory, and search ranking.
  • Read-mostly journeys; writes use sandbox payment mode/void afterwards; auto-clean created data; idempotency keys.
  • No load tests against prod unless planned/throttled; rate-limit synthetics; exclude from SLO if you must, but include their pass/fail in SLI.
  • Dark launch/shadow traffic: mirror requests to new version, compare responses/latency without user impact (goreplay, service mesh mirroring).
  • Feature flags + canary (release-readiness-testing, configuration-feature-flag-testing).

SLIs, SLOs & alerts

Define SLIs from user experience: availability (good/total requests), latency (p95 < X), correctness/freshness. SLO e.g. 99.9%/30 d → error budget 43 min. Alert on burn rate (multi-window: 2%/1 h fast, 5%/6 h slow), not raw thresholds. Test your alerts: deliberately trigger (kill synthetic, inject error via flag) and confirm page arrives, routes to right team, links runbook, and clears on recovery; test on-call escalation quarterly.

RUM (real-user monitoring)

Web vitals by route/device/geo (web-performance-testing), JS errors (Sentry), API failures per release, funnel drop-offs; segment by version to detect regressions within minutes.

Feedback loop into QA

Every incident → (1) regression test, (2) new/tuned monitor, (3) runbook fix. Track: detection time, false-alarm rate, coverage of top-10 user journeys by synthetics.

Related

observability-testing, smoke-sanity-testing, chaos-resilience-testing, release-readiness-testing, e2e-testing

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.