Where sites and services live. Choose the host that matches the workload, then layer Cloudflare and DNS on top.
Pages
- github-pages: Static site deployment with custom domains.
- vercel: Project setup, env vars, edge vs node runtimes.
- hostinger-vps: Hardening, deploys, reverse proxy.
- postgres-prod: Managed vs self-hosted, backups, PITR, failover.
- llmops-best-practices: Pillar for running LLM features in production: versioned prompts, eval gates, observability, cost control, safety, rollback.
- llm-evaluation-in-production: Golden sets, offline eval gates in CI, online quality metrics, LLM-as-judge calibration, catching model-update regressions.
- llm-observability: What to trace, log, and alert on for LLM systems: per-request traces, token and cost metrics, quality signals, tool-call spans.
- cloudflare: DNS, proxying, page rules, Workers.
- cloudflare-workers: V8 isolates, fetch handlers, bindings, wrangler, CPU and size limits.
- cloudflare-kv: Eventually consistent key-value at the edge; kv-is-a-cache rules.
- cloudflare-r2: S3-compatible object storage with zero egress; presigned URLs and lifecycle rules.
- cloudflare-durable-objects: Single-instance stateful actors, WebSockets, hibernation, alarms.
- cloudflare-cache-rules: Cache key, tiered caching, bypass logic, cf-cache-status, purging.
- cloudflare-security-headers: Transform Rules for CSP, HSTS, Referrer-Policy, Permissions-Policy.
- cloudflare-dns: Proxied vs DNS-only, CNAME flattening at the apex, DNSSEC, TTL gotchas.
- namecheap-dns: Namecheap DNS record patterns.
- secrets-and-env: Secrets and env-var hygiene: NEXT_PUBLIC_/VITE_ are client-visible, never commit .env, scan with gitleaks, rotate on exposure.
- ci-cd: CI/CD pipeline standard: the lint, typecheck, test, build, deploy gate order, GitHub Actions, caching, and required checks before merge.
- error-tracking: Application error tracking with Sentry: source-map upload, release tagging, environment separation, PII scrubbing, and alert routing.
- incident-response: Severity levels, on-call and escalation, the first ten minutes of an outage, status comms, and the blameless postmortem.
- disaster-recovery: RPO/RTO, what to back up, automated backups (Supabase PITR, pg_dump), restore drills, and offsite copies.
- pre-launch-checklist: The security and readiness twin of the launch runbook: day-one breakage, RLS live, rate limiting, auth, and legal docs.