The Ship Stack.

What it is A JavaScript runtime, package manager, script runner, and test runner.

Why I use it It is extremely fast, compact, edge-friendly, and combines several useful JavaScript utilities into one package.

What it is The standard server-side JavaScript runtime and compatibility layer.

Why I use it It remains the JavaScript standard across much of the ecosystem and covers every compatibility path where Bun cannot.

What it is Static types for the JavaScript used throughout every product.

Why I use it It is the only sensible choice for agentic engineering because richer types give coding agents more context and make their changes safer.

What it is The React framework behind the main web applications and server APIs.

Why I use it It is the full-stack React standard and keeps shipping at the edge of what production web applications can do.

What it is A content-focused web framework that ships very little browser JavaScript.

Why I use it It is a fast, simple static-site builder that still lets me use React exactly where the site needs it.

What it is The component system behind the web product interfaces.

Why I use it It is the interface standard, with the ecosystem and reusable patterns needed across every product.

What it is React Native and Expo power DiveJourney on iOS and Android.

Why I use it One React-based codebase reaches both iOS and Android without maintaining two native applications.

What it is A utility-first styling system used by the main web applications.

Why I use it It is the styling standard and its explicit utility classes are especially easy for coding agents to understand and edit safely.

What it is Source-owned interface components built on accessible primitives.

Why I use it It is a fantastic standard component foundation with accessible primitives, strong defaults, and source code I still own.

What it is A client-side server-state cache for fetching, mutations, retries, and invalidation.

Why I use it It makes data-heavy screens responsive without rebuilding cache behavior in every component.

What it is The hosted PostgreSQL, authentication, storage, and local development platform behind the products.

Why I use it It combines a fast managed backend, PostgreSQL, authentication, APIs, realtime updates, storage, and Row Level Security in one platform.

What it is The relational database that holds product, account, billing, and operational data.

Why I use it It is the relational database standard, it powers Supabase, and it keeps the underlying data portable.

What it is A typed TypeScript query and schema layer that stays close to SQL.

Why I use it It gives TypeScript applications typed schemas and queries while staying close enough to SQL to understand what the database is doing.

What it is A type-safe TypeScript query builder that maps database tables and SQL results into application types.

Why I use it It keeps complex SQL explicit while catching table, column, and result-shape mistakes in TypeScript before they reach the database.

What it is OpenAI's framework for building agents with tools, handoffs, guardrails, and traces.

Why I use it I originally ran most AI features on OpenAI models and have since diversified, but the Agents SDK remains a strong framework for structured agent workflows.

What it is A single model API that can route requests across several AI providers.

Why I use it It is a gamechanger for quickly testing models across intelligence, speed, price, and provider without rewriting the integration.

What it is Anthropic's command-line coding agent for repository work, implementation, and verification.

Why I use it I use it alongside Codex when a new Claude model is worth testing or subsidized tokens make long engineering sessions practical.

What it is The coding workspace used to inspect repositories, implement changes, and run verification.

Why I use it It is my default and the best agentic engineering environment I have found for connecting product intent to code, tests, and real browser behavior.

What it is A shared protocol for exposing product tools and data to AI clients.

Why I use it MCP is becoming the standard connection layer for a future where agents operate products and software directly.

What it is Prompt, trace, evaluation, and cost observability for AI workflows.

Why I use it LLM usage must be observed, and Langfuse is open source, generous, and purpose-built for traces, prompts, evaluations, and cost.

What it is The issue tracker connecting product decisions to implementation and release follow-up.

Why I use it It gives work clear ownership, status, and a durable next action without slowing down daily use.

What it is The workspace for longer product plans, research, and operating notes.

Why I use it It gives ideas room to develop before they become repository work or a Linear issue.

What it is The local container and Linux environment replacing Docker Desktop on my Mac.

Why I use it It is dramatically more memory-efficient than Docker Desktop and leaves more of my local machine available for coding agents and product services.

What it is The existing local container runtime used for databases and supporting product services.

Why I use it It remains installed for compatibility, but OrbStack now does the same local work with much lower memory use and faster startup.

What it is A daemonless container engine for running isolated development workloads on Linux servers.

Why I use it It is coming next for VPS coding agents because it provides Docker-compatible containers without a permanently privileged daemon.

What it is Virtual private servers for long-running coding agents, loops, and background engineering work.

Why I use it It provides strong compute value for agent workloads that need to stay alive longer than a local session or serverless function.

What it is A private mesh network connecting my Mac, phone, and VPS development environments.

Why I use it It gives long-running coding agents and development services stable private access from anywhere without exposing them to the public internet.

What it is The SSH client I use to reach VPS development environments from desktop and mobile.

Why I use it It lets me review, restart, and continue long-running agent work from my phone when I am away from the Mac.

What it is Hosting, preview deployments, server functions, and scheduled routes for the web products.

Why I use it It is the best fit for Next.js and makes previewing, deploying, and shipping web products extremely fast.

What it is Managed Redis, rate limiting, and queue-backed delivery for serverless applications.

Why I use it Its serverless Redis, queues, scheduling, and rate limiting cover the coordination jobs that do not belong in PostgreSQL.

What it is Checkout, subscriptions, billing state, and payment webhooks.

Why I use it It is the strongest all-in-one provider for payments, billing, subscriptions, webhooks, and customer self-service.

What it is Transactional email delivery for product and account messages.

Why I use it It is a fantastic developer-first email provider with a clean API that powers transactional email across the products.

What it is The alert destination for operational events, product feedback, and workflows that need human attention.

Why I use it It puts urgent system signals where I will actually see and act on them instead of creating another monitoring inbox.

What it is Feature flags and staged rollout controls.

Why I use it It replaces environment-variable flags with proper targeting, staged rollouts, instant kill switches, and safer rollback control.

What it is Consent management for analytics, privacy choices, and regional requirements.

Why I use it It is a fast way to keep consent handling consistent and GDPR-aware without rebuilding a compliance system in every product.

What it is Application errors, traces, releases, and performance context.

Why I use it It is the standard for error tracking and connects failures to the exact release, route, trace, and surrounding events.

What it is Product events, funnels, feature usage, and behavioral analytics.

Why I use it It does a bit of everything, but its product analytics, funnels, feature flags, and session evidence make it the best analytics fit for the applications.

What it is Lightweight, privacy-focused traffic analytics for this site.

Why I use it I am trying it on the static personal site because it is lightweight, privacy-focused, and avoids the heavier analytics stack.

What it is A vendor-neutral standard for application traces, metrics, and logs.

Why I use it It is the open standard that lets all observability signals share one instrumentation layer while the backends can change.

What it is Operational log storage and querying for production systems.

Why I use it It makes traces and logs fast to search without operating a separate ingestion and storage stack.

What it is External uptime checks, heartbeats, incident alerts, and status monitoring.

Why I use it ShipFoundry uses it to prove scheduled and production systems are still alive from outside the application, and the monitoring experience is excellent.

What it is Real-browser automation for product flows, screenshots, accessibility, and smoke tests.

Why I use it It is the browser-automation standard and proves the complete user flow instead of only testing isolated functions.

What it is Repeatable browser audits for performance, accessibility, and page quality.

Why I use it It turns performance into a release check when runtimes and frameworks change.

What it is The accessibility rules engine used inside automated browser tests.

Why I use it It is the standard automated check for catching common WCAG failures before they become repeated component patterns.

What it is A fast Vite-native unit and integration test runner with a mature testing ecosystem.

Why I use it It provides strong mocking, coverage, browser-like environments, and compatibility for the shared starter test suite.

What it is The unit test runner built directly into Bun.

Why I use it It starts quickly, needs almost no extra configuration, and keeps everyday tests inside the same runtime used for scripts and packages.

What it is A fast parser, linter, formatter, and TypeScript-aware toolchain.

Why I use it It is unbelievably faster than the ESLint and Prettier toolchain it replaces. The difference is large enough to change how often linting and formatting can run.

What it is The existing JavaScript and TypeScript linter in repositories not yet moved to Oxlint.

Why I use it It remains only while its rules are transferred without weakening the checks.

What it is The formatter still used where Oxfmt has not yet replaced it.

Why I use it It stays until the new formatter can produce stable output without noisy repository-wide changes.

What it is Continuous integration, release automation, and scheduled repository work.

Why I use it Checks should run beside the pull request and commit they are evaluating.

What it is Automated pull request review for bugs, regressions, and code-quality problems.

Why I use it It adds a consistent review pass before a human spends time on the final decision.

What it is Repository-aware review of the implementation, tests, product behavior, and release risk.

Why I use it It can connect a code diff to the surrounding system instead of reviewing changed lines in isolation.

What it is High-performance hosted machines used for selected GitHub Actions jobs.

Why I use it They shorten expensive build and browser workloads when GitHub-hosted machines become the bottleneck.

What it is An external runner lane used during the fleet workflow experiments.

Why I use it It is tracked because runner choice affects delivery speed, reliability, and trust boundaries.

What it is Managed GitHub Actions runners used for selected small and medium workloads.

Why I use it They provide another cost and performance option while trusted jobs remain explicitly routed.

What it is The structured-content platform and editing studio behind johnpotess.com.

Why I use it Essays, metadata, translations, and related content need a real content model rather than loose files.

What it is The documentation framework used for ShipFoundry knowledge and product guidance.

Why I use it It provides a strong public reading and search experience without rebuilding a documentation system.

What it is Web and native maps, place search, and geographic imagery for DiveJourney.

Why I use it The product needs one geographic foundation across discovery, planning, and the mobile experience.