monitoring

Rules for monitoring software services and infrastructure using measurements, service targets, and alerts. An SLO is a target for how reliably or quickly a service should work for users.

In plain words
What is it for?
Use it to define latency, traffic, error, and resource measurements; set availability or latency targets; and design actionable alerts.
Why use it?
It helps teams notice user-facing problems, avoid noisy alerts, and connect each alert to a response procedure.

Cursor rule

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/nedcodes-ok/cursor-doctor/monitoring
Clone the repo
git clone --depth 1 https://github.com/nedcodes-ok/cursor-doctor
Per session 513 This file is loaded in full into every session.
When invoked 513 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00513 $0.00513
Opus 5 $0.00257 $0.00257
Sonnet 5 $0.00103 $0.00103
Haiku 4.5 $0.00051 $0.00051

Measured 2d ago against content hash 2b42063f37a9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

monitoring scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

pro-kit/templates/practices/monitoring.mdc · 51 lines

How it starts

The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Monitoring Cursor Rules

You are an expert at application monitoring. Follow these rules:

Four Golden Signals

  • Latency: track P50, P95, P99 — averages hide tail latency
  • Traffic: requests per second by endpoint and status code
  • Errors: error rate as percentage, broken down by type
  • Saturation: CPU, memory, disk, connection pool utilization

Metrics

  • USE method for infrastructure: Utilization, Saturation, Errors
  • RED method for services: Rate, Errors, Duration
  • Custom business metrics: signups/min, orders/hour, revenue
  • Use histograms for latency, not averages — P99 matters most
  • Label dimensions: service, endpoint, status_code, environment

Alerting

  • Alert on symptoms (high error rate), not causes (CPU spike)
  • Two tiers: page (P1, needs human now) and notify (P2, next business day)
  • Burn-rate alerts for SLO-based monitoring — catches slow degradation
  • Every alert needs a runbook link with triage steps
  • No alert without an actionable response — remove noisy alerts ruthlessly

SLOs

  • Define SLIs first: what indicates the service is working for users
  • Availability SLO: 99.9% = 43 min downtime/month — pick realistic targets
  • Latency SLO: 95% of requests under 200ms, 99% under 1s
  • Error budget: when budget is exhausted, freeze features and fix reliability
  • Review SLOs quarterly — adjust based on actual user impact

Dashboards

  • Service overview: golden signals at a glance
  • Dependency dashboard: health of all downstream services
  • Business dashboard: key metrics non-engineers care about
  • No dashboard with more than 10 panels — if everything is important, nothing is
  • Include links from dashboards to relevant logs and traces

Health Checks

  • /health for load balancers: returns 200 if the process is running
  • /ready for k8s readiness: returns 200 only when ready to serve traffic
  • Deep health checks: verify DB, cache, external service connectivity
  • Dont alert on brief health check failures — use consecutive failure thresholds
  • Synthetic monitoring: probe critical user flows every 1-5 minutes

Read the full file on GitHub · 51 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 51 lines · 513 tokens per session scan A 2b42063f37a9

Subscribe to this mod's changes

monitoring is a cursor rule published in the GitHub repository nedcodes-ok/cursor-doctor (9 stars, last pushed 5mo ago), licensed MIT. It adds 513 tokens to every session, about $0.0026 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.