Update Policy

Purpose

This policy keeps the guide current without chasing every trend. Updates should improve decision quality, not merely add new tools.

Review cadence

Area	Cadence	What to check
React and rendering	Quarterly	RSC, compiler, framework caching, streaming, server actions, hydration guidance.
Web performance	Quarterly	Web Vitals definitions, thresholds, measurement guidance, browser changes.
Accessibility	Twice yearly	WCAG updates, browser/assistive technology guidance, component pattern changes.
Browser security/privacy	Quarterly	CSP, Trusted Types, cookies, storage, passkeys, CORS, privacy changes.
AI/GenUI/MCP/A2A	Monthly while volatile	Protocol changes, SDK behavior, security concerns, eval practices, production patterns.
Local source-derived content	Twice yearly	Gaps, stale examples, shallow chapters, missing exercises.

Update workflow

1. Identify the chapter and source dependency.
2. Verify current primary sources for unstable topics.
3. Decide whether the change affects advice, examples, checklists, diagrams, or rubrics.
4. Update source lens or source-map notes when source dependency changes.
5. Run build and content checks.
6. Add backlog items for follow-up rather than bloating the current change.

Material change threshold

Update a chapter when one of these is true:

• A platform API, framework behavior, protocol, or metric changed.
• Existing guidance could lead to production risk.
• A chapter lacks enough decision criteria, failure modes, or verification steps.
• A capstone no longer reflects realistic production constraints.
• A review packet misses a common failure mode seen in incidents or modern practice.

Non-goals

• Do not add every new framework.
• Do not rewrite chapters for style churn only.
• Do not make AI topics sound stable when the ecosystem is still shifting.
• Do not replace durable architecture principles with vendor-specific recipes.