Skip to main content

Frontend Architect Foundation Map

Why this chapter matters

Part 0 is the base layer for the whole book. A frontend architect does not only know React, CSS, or build tools. They understand how product intent becomes browser behavior under real constraints: latency, device limits, accessibility, security, data freshness, ownership, release risk, and team scale.

This chapter gives you the map. The rest of Part 0 fills in the operating skills.

The baseline expectation

Before architect-level system design, you should be able to reason across ten foundations:

FoundationArchitect-level question
Product and user intentWhat user outcome is protected, and what failure would users actually feel?
Browser runtimeWhat work happens on the main thread, network, storage, compositor, and accessibility tree?
Rendering and dataWhich state belongs in URL, server cache, client cache, local storage, or component memory?
Interaction designWhat happens across loading, empty, error, partial, disabled, pending, and recovery states?
AccessibilityCan the workflow be completed with keyboard, screen reader, zoom, reduced motion, and constrained cognition?
Security and privacyWhat input, script, storage, identity, and third-party boundaries are trusted or untrusted?
PerformanceWhat is the route-level cost in bytes, CPU, network, rendering, and user-perceived latency?
VerificationWhich tests, budgets, audits, traces, and production signals prove the behavior survives release?
DeliveryHow do build, deploy, rollback, flags, experiments, and dependencies change risk?
OwnershipWho can change the surface, who approves exceptions, and who responds when it fails?

If one of these is missing, the architecture decision is probably incomplete.

The frontend architect's mental stack

Business goal
-> user journey
-> interaction contract
-> data and permission model
-> rendering and runtime strategy
-> component and module boundaries
-> quality gates
-> release and rollback plan
-> production telemetry
-> ownership and evolution

Framework features enter late in this stack. They matter, but they implement decisions rather than replace them.

The four contracts of a frontend surface

Every important frontend surface has four contracts.

1. User contract

The user contract defines what the user can expect.

  • What can they accomplish?
  • What state is visible?
  • What can fail?
  • How do they recover?
  • What works on slow devices, small screens, assistive technology, or poor networks?

2. Data contract

The data contract defines truth and freshness.

  • Who owns the data?
  • How fresh must it be?
  • What happens when reads are stale?
  • What happens when mutations fail?
  • How are permissions enforced?
  • What data can be cached, logged, or stored locally?

3. Runtime contract

The runtime contract defines browser and delivery behavior.

  • What renders on the server, edge, or client?
  • What JavaScript must run before the user can act?
  • What code is shared across routes?
  • Which APIs, workers, caches, and storage layers are used?
  • How does the surface degrade?

4. Team contract

The team contract defines change.

  • Who owns the route, module, API, component, and quality gates?
  • What changes require review?
  • What standards apply?
  • What exceptions exist?
  • When is debt revisited?

Foundation maturity rubric

LevelBehavior
DevelopingCan implement features but mostly reasons from component code and framework examples.
Strong seniorCan reason about route behavior, state ownership, accessibility, performance, and production failure.
Staff-readyCan define contracts, tradeoffs, quality gates, and migration paths across multiple teams or surfaces.
Architect-readyCan create decision systems that make other teams faster and safer without becoming the bottleneck.

Minimum artifact set

A frontend architect should be comfortable producing these artifacts before moving into later parts of the book:

ArtifactPurpose
Route architecture noteRendering mode, data dependencies, cache/freshness, failure states, budgets
Interaction state matrixLoading, empty, error, pending, offline, unauthorized, partial success
State ownership mapURL, server, client cache, local durable storage, component state
Accessibility contractKeyboard, focus, semantics, names, live regions, reduced motion
Threat boundary sketchUser input, third-party scripts, storage, auth/session, dangerous sinks
Performance budgetRoute bytes, LCP, INP, CLS, expensive interactions, monitoring
Verification planTests, checks, manual review, field metrics, rollback signal
Ownership registerOwner, approver, exception, expiry, escalation path

How to use Part 0

Read Part 0 in this order if you want a broad foundation:

  1. Foundation map and senior mindset.
  2. Performance, browser runtime, forms, security, testing, tooling, and observability.
  3. Domain/data/state modeling and production readiness.
  4. Weakness diagnosis, transition mistakes, debugging drills, ownership exercises, and architect thinking.
  5. Advanced platform frontiers after the baseline is comfortable.

Do not rush to framework-specific architecture before you can explain the route, data, runtime, user, and ownership contracts without naming a framework.

Exercises

  1. Pick one production route and fill out the four contracts: user, data, runtime, and team.
  2. Create the minimum artifact set for one feature you already know well.
  3. Mark each foundation area as weak, usable, strong, or teachable.
  4. Choose one weak area and define a two-week practice loop with a concrete output.