web-perf
Deep Core Web Vitals and page speed audit using Chrome DevTools MCP.
Skill by PanPanFR
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/panpanfr-web-perf
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: security, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/web-perf
What's inside
- Skill name
web-perf- Description
- Deep Core Web Vitals and page speed audit using Chrome DevTools MCP. Measures LCP, INP, CLS, FCP, TTFB, TBT, Speed Index. Identifies render-blocking resources, network chains, layout shifts, caching issues, accessibility gaps. Use for performance audit, optimization, Lighthouse scores, site speed. Biases towards retrieval from current docs.
- License
- MIT
- Instructions
- 160 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Web Performance Audit
Unified skill combining audit-speed thresholds with Chrome DevTools MCP workflows.
## Core Web Vitals Thresholds
| Metric | Good | Needs Improvement | Poor |
|--------|------|-------------------|------|
| LCP | < 2.5s | 2.5s - 4.0s | > 4.0s |
| CLS | < 0.1 | 0.1 - 0.25 | > 0.25 |
| INP | < 200ms | 200ms - 500ms | > 500ms |
| FCP | < 1.8s | 1.8s - 3.0s | > 3.0s |
| TTFB | < 800ms | 800ms - 1800ms | > 1800ms |
| TBT | < 200ms | 200ms - 600ms | > 600ms |
| Speed Index | < 3.4s | 3.4s - 5.8s | > 5.8s |
## Retrieval Sources
| Source | URL | Use for |
|--------|-----|---------|
| web.dev | https://web.dev/articles/vitals | CWV thresholds, definitions |
| Chrome DevTools | https://developer.chrome.com/docs/devtools/performance | Tooling APIs, trace analysis |
| Lighthouse | https://developer.chrome.com/docs/lighthouse/performance/performance-scoring | Score weights, metric thresholds |
## MCP Setup Required
Add to your MCP config:
```json
"chrome-devtools": {
"type": "local",
"command": ["npx", "-y", "chrome-devtools-mcp@latest"]
}
```
## Workflow
### Phase 1: Performance Trace
```
navigate_page(url: "<target-url>")
performance_start_trace(autoStop: true, reload: true)
```
### Phase 2: Core Web Vitals Analysis
Use `performance_analyze_insight` with these insight names:
| Metric | Insight Name | What to Look For |
|--------|--------------|------------------|
| LCP | LCPBreakdown | TTFB, resource load, render delay breakdown |
| CLS | CLSCulprits | Elements causing shifts (no dims, font swap, injected) |
| INP | INPBreakdown | Main thread blocking, long tasks, event handlers |
| Render Blocking | RenderBlocking | CSS/JS blocking first paint |
| Network Dependencies | NetworkRequestsDepGraph | Request chains delaying critical resources |
### Phase 3: Network Analysis
```
list_network_requests(resourceTypes: ["Script", "Stylesheet", "Document", "Font", "Image"])
```
Look for:
1. Render-blocking resources in `<head>` without async/defer
2. Network chains (late-discovered resources)
3. Missing preloads (critical fonts, hero images, key scripts)
4. Caching issues (missing/weak Cache-Control, ETag)
5. Large payloads (uncompressed/oversized bundles)
6. Unused preconnects (zero requests to that origin)
### Phase 4: Accessibility Snapshot
```
take_snapshot(verbose: true)
```
Flag: missing/duplicate ARIA IDs, poor contrast (WCAG AA: 4.5:1), focus traps, missing accessible names
### Phase 5: Codebase Analysis (if access)
Detect framework/bundler:
| Tool | Config Files |
|------|--------------|
| Webpack | webpack.config.js, webpack.*.js |
| Vite | vite.config.js, vite.config.ts |
| Next.js | next.config.js, next.config.mjs |
| Nuxt | nuxt.config.js, nuxt.config.ts |
| Astro | astro.config.mjs |
Check for:
- Tree-shaking config (Webpack: sideEffects, usedExports; Vite/Rollup: default on)
- Unused CSS/JS (Tailwind content config, PurgeCSS, dynamic imports)
- Polyfills (@babel/preset-env targets, core-js size)
- Compression (terser/esbuild/swc, gzip/brotli, source maps in prod)
## LCP Root-Cause Tree
**TTFB slow (>800ms)?** → Server: hosting, CDN, DB queries, SSR time → Fix: upgrade hosting, CDN, optimize server, enable caching
**LCP element is image?** → Image: format (WebP/AVIF), size (responsive), lazy loading (NO on LCP) → Fix: modern formats, width/height, fetchpriority="high", preload
**LCP element is text?** → Font: custom fonts blocking, file size, font-display → Fix: font-display: swap/optional, preload critical, subset
**Render-blocking CSS/JS?** → Large CSS in `<head>`, sync JS before content → Fix: inline critical CSS, defer non-critical, async/defer JS
## CLS Root-Cause Tree
**Images/videos lack dimensions?** → Fix: width/height attributes, CSS aspect-ratio
**Ads/embeds inject content?** → Fix: reserve space with min-height, contain-intrinsic-size
**Fonts cause reflow?** → Fix: font-display: optional, or match fallback metrics
**Dynamic content inserts above fold?** → Fix: use overlays, reserve space with fixed-height containers
## INP Root-Cause Tree
**Main thread blocked by long tasks?** → JS execution, third-party, large DOM → Fix: break tasks with requestIdleCallback/setTimeout, code-split
**Event handlers do heavy sync work?** → Click handlers with large DOM updates, validation on keystroke → Fix: debounce, requestAnimationFrame, web workers
**Third-party scripts competing?** → Analytics, chat, A/B testing, social embeds → Fix: defer until after interaction, loading="lazy", remove low-value
## Resource Analysis Table
| Resource Type | Size | Assessment | Action |
|---|---|---|---|
| HTML | [x] KB | [ok/large] | Compress, reduce inline |From PanPanFR/oh-my-kilo, MIT.
More from PanPanFR/oh-my-kilo
- 0.594score
caveman
PanPanFR
Ultra-compressed communication mode.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
clean-code
PanPanFR
This skill embodies the principles of "Clean Code" by Robert C.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
documentation
PanPanFR
Write and maintain technical documentation.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
ponytail-audit
PanPanFR
Whole-repo audit for over-engineering.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
ponytail-review
PanPanFR
Code review focused exclusively on over-engineering.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
ui-design
PanPanFR
UI/UX design intelligence for web and mobile.
SkillBFreshness: B★ 12Claude Code +9 - 0.591score
ponytail
PanPanFR
Lazy senior dev mode.
SkillBFreshness: B★ 12Claude Code +9 - 0.569score
code-review
PanPanFR
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding…
SkillBFreshness: B★ 12Claude Code +9 - 0.569score
codebase-design
PanPanFR
Shared vocabulary for designing deep modules.
SkillBFreshness: B★ 12Claude Code +9 - 0.568score
executing-plans
PanPanFR
Use when you have a written implementation plan to execute in a separate session with review checkpoints
SkillBFreshness: B★ 12Claude Code +9 - 0.568score
writing-plans
PanPanFR
Use when you have a spec or requirements for a multi-step task, before touching code
SkillBFreshness: B★ 12Claude Code +9 - 0.325score
cloudflare
PanPanFR
No description from the publisher yet.
SkillBFreshness: B★ 12Claude Code +9
Signals
- Stars
- 12
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Aug 10, 2026 (2 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/PanPanFR/oh-my-kilo
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from PanPanFR
- 0.594score
caveman
PanPanFR
Ultra-compressed communication mode.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
clean-code
PanPanFR
This skill embodies the principles of "Clean Code" by Robert C.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
documentation
PanPanFR
Write and maintain technical documentation.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
ponytail-audit
PanPanFR
Whole-repo audit for over-engineering.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
ponytail-review
PanPanFR
Code review focused exclusively on over-engineering.
SkillBFreshness: B★ 12Claude Code +9 - 0.594score
ui-design
PanPanFR
UI/UX design intelligence for web and mobile.
SkillBFreshness: B★ 12Claude Code +9
Cite this page
web-perf skill. Agent Mart. https://agentmart.dev/skills/panpanfr-web-perf (last updated 2026-10-24)