Quick Reference

Quick reference for common tasks, commands, and endpoints.

Endpoints

AI & LLM

URL	Purpose	For
`/llms.txt`	Discovery file	AI agents
`/llms-full.txt`	All docs, structured format	RAG systems
`/docs/page.mdx`	Page as markdown	AI tools
`/docs/page.md`	Page as markdown (alt)	AI tools

/llms-full.txt has enhanced structure with metadata, TOC, and clear separators. See Format Documentation.

RSS Feeds

URL	Format	Content
`/rss.xml`	RSS 2.0	Sitewide
`/atom.xml`	Atom 1.0	Sitewide
`/feed.json`	JSON Feed	Sitewide
`/docs/rss.xml`	RSS 2.0	Docs only
`/docs/atom.xml`	Atom 1.0	Docs only
`/docs/feed.json`	JSON Feed	Docs only

All feeds are auto-discoverable via <link> tags and refreshed hourly. See RSS Feeds.

Content Negotiation

# Get markdown automatically
curl -H "Accept: text/markdown" https://yourdomain.com/docs/page

Scripts

Development

# Start dev server
pnpm dev

# Start on different port
PORT=3001 pnpm dev

PDF Export

# Export all pages
pnpm export-pdf

# Export specific pages
pnpm export-pdf:specific /docs /docs/getting-started

# Build and export (production)
pnpm export-pdf:build

Link Validation

# Validate all documentation links
pnpm lint:links

Build & Deploy

# Build for production
pnpm build

# Start production server
pnpm start

# Build with PDF export mode
NEXT_PUBLIC_PDF_EXPORT=true pnpm build

Testing Commands

Test Endpoints

# Discovery file
curl http://localhost:3000/llms.txt

# Full documentation
curl http://localhost:3000/llms-full.txt

# Specific page as markdown
curl http://localhost:3000/docs.mdx

# Test content negotiation
curl -H "Accept: text/markdown" http://localhost:3000/docs

Inspect Headers

# Check HTTP headers
curl -I http://localhost:3000/llms.txt
curl -I http://localhost:3000/llms-full.txt

# Check custom headers
curl -I http://localhost:3000/llms-full.txt | grep "X-"

Download Content

# Download full documentation
curl http://localhost:3000/llms-full.txt -o docs.txt

# View table of contents
curl http://localhost:3000/llms-full.txt | \
  sed -n '/Table of Contents:/,/^===/p'

# Count pages
curl http://localhost:3000/llms-full.txt | \
  grep -c "^PAGE [0-9]"

Components

AI Page Actions

import { LLMCopyButton, ViewOptions } from '@/components/page-actions';

// Copy markdown button
<LLMCopyButton markdownUrl={`${page.url}.mdx`} />

// View options dropdown
<ViewOptions
  markdownUrl={`${page.url}.mdx`}
  githubUrl="https://github.com/user/repo/blob/main/..."
/>

Cards and Callouts

import { Card, Cards } from "fumadocs-ui/components/card";
import { Callout } from "fumadocs-ui/components/callout";

<Cards>
  <Card title="Title" description="Description" href="/link" />
</Cards>

<Callout type="info">Information message</Callout>

<Callout type="warn">Warning message</Callout>

Tabs and Steps

import { Tab, Tabs } from "fumadocs-ui/components/tabs";
import { Step, Steps } from "fumadocs-ui/components/steps";

<Tabs items={["Tab 1", "Tab 2"]}>
  <Tab value="Tab 1">Content 1</Tab>
  <Tab value="Tab 2">Content 2</Tab>
</Tabs>

<Steps>
  <Step>First step</Step>
  <Step>Second step</Step>
</Steps>

Configuration Files

Key Files

File	Purpose
`middleware.ts`	Content negotiation
`next.config.mjs`	URL rewrites
`source.config.ts`	MDX processing
`app/llms.txt/route.ts`	Discovery endpoint
`components/page-actions.tsx`	AI UI components

Environment Variables

# Enable PDF export mode
NEXT_PUBLIC_PDF_EXPORT=true

# Set port
PORT=3001

Customization

Update GitHub URL

Edit app/docs/[[...slug]]/page.tsx:

githubUrl={`https://github.com/YOUR_ORG/YOUR_REPO/blob/main/apps/web/content/docs/${page.file.path}`}

Add AI Tool

Edit components/page-actions.tsx:

{
  title: 'Open in MyAI',
  href: `https://myai.com/?url=${markdownUrl}`,
  icon: <MyIcon />,
}

Customize PDF Settings

Edit scripts/export-pdf.ts:

await page.pdf({
  width: "950px",
  margin: { top: "20px", right: "20px", bottom: "20px", left: "20px" },
});

Modify Discovery File

Edit app/llms.txt/route.ts:

const content = `# Your Custom Title

> Your custom description

## Documentation Pages

${pages.map((page) => `- [${page.data.title}](${origin}${page.url}.mdx): ${page.data.description ?? ""}`).join("\n")}
`;

Common Tasks

Add New Doc Page

Create file in content/docs/
Add frontmatter with title and description
Write MDX content
Page automatically appears in navigation

Export PDFs

Start dev server: pnpm dev
Run export: pnpm export-pdf
Find PDFs in pdfs/ directory

Test AI Integration

Start dev server: pnpm dev
Test endpoints with curl (see above)
Check page actions in browser
Verify content negotiation

Deploy to Production

Build: pnpm build
Test locally: pnpm start
Deploy to hosting platform
Verify all endpoints work

Performance Optimization

Caching Strategy

Resource	Cache-Control
`/llms.txt`	`s-maxage=86400` (24h)
`/llms-full.txt`	`revalidate=false` (permanent)
`*.mdx` routes	`immutable` (forever)

Build Optimization

# Analyze bundle
pnpm build --analyze

# Check bundle size
du -sh .next/

# Clear cache
rm -rf .next/

Troubleshooting

Clear Cache

# Clear Next.js cache
rm -rf .next/

# Clear node modules
rm -rf node_modules/
pnpm install

Check Errors

# View build errors
pnpm build 2>&1 | tee build.log

# Check TypeScript errors
pnpm tsc --noEmit

Verify Configuration

# Check source config
cat source.config.ts | grep includeProcessedMarkdown

# Check middleware
cat middleware.ts

# Check rewrites
cat next.config.mjs | grep -A 10 "rewrites"

Documentation

Full Guides

Getting Started - Overview and introduction
PDF Export - Complete PDF export guide
AI Integration - AI/LLM integration guide
llms-full.txt Format - Format specification
Testing Guide - Comprehensive testing

External Resources

Status

✅ All features implemented and tested ✅ Following Fumadocs official guidelines ✅ Performance optimized with caching ✅ Production ready

Quick Reference

On this page