name: cwv-improvement-planner description: Build and prioritize Core Web Vitals improvement plans for web apps with a focus on LCP, INP, and TTFB. Use when users mention CWV, Lighthouse, PageSpeed, performance budgets, render blocking, hydration, image prioritization, CDN caching/compression, CloudFront, or Next.js web performance tuning.
CWV Improvement Planner
Goal
Turn performance findings into a prioritized, measurable rollout plan that improves field Core Web Vitals without platform-specific assumptions.
Use This Skill When
- The user asks for a CWV plan, optimization roadmap, or performance triage.
- The context includes LCP, INP, CLS, TTFB, render-blocking resources, hydration, or preload strategy.
- The stack mentions Next.js, React, CDN/edge caching (including CloudFront), or page-speed regressions.
Cross-Platform Rules
- Use forward-slash paths only.
- Avoid OS-specific command syntax unless the repository already standardizes it.
- Keep recommendations tool-agnostic first, then add stack-specific examples when context confirms the stack.
Workflow
- Establish baseline
- Capture field metrics (RUM/CrUX) and lab traces (Lighthouse/WebPageTest/DevTools).
- Identify highest-impact templates and user journeys (for example, list -> detail -> checkout).
- Prioritize fundamentals first
- Compression and caching correctness at the edge.
- TTFB contributors and redirect chains.
- LCP resource discovery, eager loading, and prioritization.
- Render-blocking CSS/JS and hydration pressure.
- Convert findings into implementation slices
- For each issue: why it matters, what to inspect, recommended change, risk, estimated impact.
- Add acceptance criteria and rollback notes.
- Gate advanced experiments
- Treat 103 Early Hints and Speculation Rules as measured experiments with explicit guardrails.
- Define verification
- Add before/after checks in CI and post-release monitoring thresholds.
Output Template
Use this structure when generating a plan:
# Core Web Vitals Improvement Plan
## Baseline
- Field baseline: LCP, INP, CLS, TTFB by device/network/route.
- Lab baseline: trace links and key waterfall observations.
## Prioritized Work
1. [High] <topic>
- Why
- Inspect
- Change
- Risks
- Expected impact
2. [Medium] <topic>
...
## Experiments
- Hypothesis
- Scope
- Success metric
- Stop conditions
## Validation
- CI assertions
- Release checks
- RUM watch window and thresholds
Detailed Guidance
See playbook.md for the topic-by-topic checklist and prioritization model.