search-indexing

name	search-indexing
description	Use when configuring, updating, or debugging the Algolia/Pinecone search indexing pipeline for documentation

Search Indexing Patterns

Overview

Manage search indexing with Algolia for full-text search and Pinecone for vector search. Index updates run via npm script after build.

When to Use

• Updating search index after content changes
• Debugging search results
• Adding new search features
• Configuring search settings
• Troubleshooting missing content in search

Indexing Pipeline

digraph indexing {
  "MDX Content" -> "static.json" [label="build"]
  "static.json" -> "Algolia" [label="npm run index"]
}

Commands

npm run index    # Update Algolia index
npm run build    # Build + update index

Key Files

File	Purpose
scripts/update-index.mjs	Index sync script
lib/algolia-client.ts	Algolia search client
lib/pinecone.ts	Pinecone vector client
app/api/search/route.ts	Search API endpoint

Algolia Configuration

// Index settings in update-index.mjs
{
  queryLanguages: ['en', 'th'],
  searchableAttributes: ['title', 'description', 'content', 'structured.headings'],
  customRanking: ['desc(weight.title)', 'desc(weight.content)'],
}

Environment Variables

NEXT_PUBLIC_ALGOLIA_APP_ID=
NEXT_PUBLIC_ALGOLIA_SEARCH_API_KEY=
ALGOLIA_ADMIN_API_KEY=          # server-side only, no NEXT_PUBLIC_ prefix
PINECONE_API_KEY=
PINECONE_INDEX_NAME=

Common Issues

• Missing content in search: Run npm run index after adding new docs
• Stale results: Clear browser cache or re-run index
• Missing env vars: Create .env.local with required keys

Quick Reference

Task	Command
Update index	npm run index
Full rebuild	npm run build
Test search	Visit /docs and use search