Mapbox Web Integration Patterns logo

Mapbox Web Integration Patterns

Organization
mapbox
mapbox-web-integration-patterns

Official integration patterns for Mapbox GL JS across popular web frameworks (React, Vue, Svelte, Angular). Covers setup, lifecycle management, token handling, search integration, and common pitfalls. Based on Mapbox's create-web-app scaffolding tool.

Overview

Publishermapbox
Repositorymapbox-agent-skills
Skill namemapbox-web-integration-patterns
Stars
78
Forks
17
Bundled files
10
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 10 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by mapbox on GitHub. Read the source before you install it.

Installation

Install the Mapbox Web Integration Patterns AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/mapbox/mapbox-agent-skills.git /tmp/mapbox-agent-skills
mkdir -p .claude/skills
cp -r /tmp/mapbox-agent-skills/skills/mapbox-web-integration-patterns .claude/skills/mapbox-web-integration-patterns
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Mapbox Web Integration Patterns in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Mapbox Web Integration Patterns on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Mapbox Web Integration Patterns is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Mapbox Integration Patterns Skill

This skill provides official patterns for integrating Mapbox GL JS into web applications using React, Vue, Svelte, Angular, and vanilla JavaScript. These patterns are based on Mapbox's create-web-app scaffolding tool and represent production-ready best practices.

Version Requirements

Mapbox GL JS

Recommended: v3.x (latest)

  • Minimum: v3.0.0
  • Why v3.x: Modern API, improved performance, active development
  • v2.x: Legacy; no longer actively developed (see migration notes below)

Installing via npm (recommended for production):

bash
npm install mapbox-gl@^3.0.0    # Installs latest v3.x

CDN (for prototyping only):

html
<!-- Replace VERSION with latest v3.x from https://docs.mapbox.com/mapbox-gl-js/ -->
<script src="https://api.mapbox.com/mapbox-gl-js/vVERSION/mapbox-gl.js"></script>
<link href="https://api.mapbox.com/mapbox-gl-js/vVERSION/mapbox-gl.css" rel="stylesheet" />

Framework Requirements

React: GL JS works with React 16.8+ (requires hooks). create-web-app scaffolds with React 19.x. Vue: GL JS works with Vue 2.x+ (Vue 3 Composition API recommended). Svelte: GL JS works with any Svelte version. create-web-app scaffolds with Svelte 5.x. Angular: GL JS works with Angular 2+. create-web-app scaffolds with Angular 19.x. Next.js: Minimum 13.x (App Router), Pages Router 12.x+.

Mapbox Search JS

bash
npm install @mapbox/search-js-react@^1.0.0      # React
npm install @mapbox/search-js-web@^1.0.0        # Other frameworks

Version Migration Notes (v2.x to v3.x)

  • WebGL 2 now required
  • optimizeForTerrain option removed
  • Improved TypeScript types, better tree-shaking support
  • No breaking changes to core initialization patterns

Token patterns (work in v2.x and v3.x):

javascript
const token = import.meta.env.VITE_MAPBOX_ACCESS_TOKEN; // Use env vars in production

// Global token (works since v1.x)
mapboxgl.accessToken = token;
const map = new mapboxgl.Map({ container: '...' });

// Per-map token (preferred for multi-map setups)
const map = new mapboxgl.Map({
  accessToken: token,
  container: '...'
});

Core Principles

Every Mapbox GL JS integration must:

  1. Initialize the map in the correct lifecycle hook
  2. Store map instance in component state (not recreate on every render)
  3. Always call map.remove() on cleanup to prevent memory leaks
  4. Handle token management securely (environment variables)
  5. Import CSS: import 'mapbox-gl/dist/mapbox-gl.css'

React Integration (Primary Pattern)

Pattern: useRef + useEffect with cleanup

Note: These examples use Vite (the bundler used in create-web-app). If using Create React App, replace import.meta.env.VITE_MAPBOX_ACCESS_TOKEN with process.env.REACT_APP_MAPBOX_TOKEN. See Token Management Patterns for other bundlers.

jsx
import { useRef, useEffect } from 'react';
import mapboxgl from 'mapbox-gl';
import 'mapbox-gl/dist/mapbox-gl.css';

function MapComponent() {
  const mapRef = useRef(null); // Store map instance
  const mapContainerRef = useRef(null); // Store DOM reference

  useEffect(() => {
    mapboxgl.accessToken = import.meta.env.VITE_MAPBOX_ACCESS_TOKEN;

    mapRef.current = new mapboxgl.Map({
      container: mapContainerRef.current,
      center: [-71.05953, 42.3629],
      zoom: 13
    });

    // CRITICAL: Cleanup to prevent memory leaks
    return () => {
      mapRef.current.remove();
    };
  }, []); // Empty dependency array = run once on mount

  return <div ref={mapContainerRef} style={{ height: '100vh' }} />;
}

Key points:

  • Use useRef for both map instance and container
  • Initialize in useEffect with empty deps []
  • Always return cleanup function that calls map.remove()
  • Never initialize map in render (causes infinite loops)

React + Search JS

jsx
import { useRef, useEffect, useState } from 'react';
import mapboxgl from 'mapbox-gl';
import { SearchBox } from '@mapbox/search-js-react';
import 'mapbox-gl/dist/mapbox-gl.css';

const accessToken = import.meta.env.VITE_MAPBOX_ACCESS_TOKEN;
const center = [-71.05953, 42.3629];

function MapWithSearch() {
  const mapRef = useRef(null);
  const mapContainerRef = useRef(null);
  const [inputValue, setInputValue] = useState('');

  useEffect(() => {
    mapboxgl.accessToken = accessToken;

    mapRef.current = new mapboxgl.Map({
      container: mapContainerRef.current,
      center: center,
      zoom: 13
    });

    return () => {
      mapRef.current.remove();
    };
  }, []);

  return (
    <>
      <div
        style={{
          margin: '10px 10px 0 0',
          width: 300,
          right: 0,
          top: 0,
          position: 'absolute',
          zIndex: 10
        }}
      >
        <SearchBox
          accessToken={accessToken}
          map={mapRef.current}
          mapboxgl={mapboxgl}
          value={inputValue}
          proximity={center}
          onChange={(d) => setInputValue(d)}
          marker
        />
      </div>
      <div ref={mapContainerRef} style={{ height: '100vh' }} />
    </>
  );
}

Search JS Integration Summary

Install:

bash
npm install @mapbox/search-js-react      # React
npm install @mapbox/search-js-web        # Vanilla/Vue/Svelte

Both packages include @mapbox/search-js-core as a dependency. Only install -core directly if building a custom search UI.

Key configuration options:

  • accessToken: Your Mapbox public token
  • map: Map instance (must be initialized first)
  • mapboxgl: The mapboxgl library reference
  • proximity: [lng, lat] to bias results geographically
  • marker: Boolean to show/hide result marker
  • placeholder: Search box placeholder text

Positioning Search Box

Absolute positioning (overlay):

jsx
<div
  style={{
    position: 'absolute',
    top: 10,
    right: 10,
    zIndex: 10,
    width: 300
  }}
>
  <SearchBox {...props} />
</div>

Common positions:

  • Top-right: top: 10px, right: 10px
  • Top-left: top: 10px, left: 10px
  • Bottom-left: bottom: 10px, left: 10px

Common Mistakes (Critical)

Mistake 1: Forgetting to call map.remove()

javascript
// BAD - Memory leak!
useEffect(() => {
  const map = new mapboxgl.Map({ ... })
  // No cleanup function
}, [])

// GOOD - Proper cleanup
useEffect(() => {
  const map = new mapboxgl.Map({ ... })
  return () => map.remove()  // Cleanup
}, [])

Why: Every Map instance creates WebGL contexts, event listeners, and DOM nodes. Without cleanup, these accumulate and cause memory leaks.

Mistake 2: Initializing map in render

javascript
// BAD - Infinite loop in React!
function MapComponent() {
  const map = new mapboxgl.Map({ ... })  // Runs on every render
  return <div />
}

// GOOD - Initialize in effect
function MapComponent() {
  useEffect(() => {
    const map = new mapboxgl.Map({ ... })
  }, [])
  return <div />
}

Why: React components re-render frequently. Creating a new map on every render causes infinite loops and crashes.

Mistake 3: Not storing map instance properly

javascript
// BAD - map variable lost between renders
function MapComponent() {
  useEffect(() => {
    let map = new mapboxgl.Map({ ... })
    // map variable is not accessible later
  }, [])
}

// GOOD - Store in useRef
function MapComponent() {
  const mapRef = useRef()
  useEffect(() => {
    mapRef.current = new mapboxgl.Map({ ... })
    // mapRef.current accessible throughout component
  }, [])
}

Why: You need to access the map instance for operations like adding layers, markers, or calling remove().

Mistake 4: Storing map instance in Vue's data() (Vue-specific)

javascript
// BAD - Vue's reactivity wraps data() objects in a Proxy, breaking mapbox-gl internals!
export default {
  data() {
    return {
      map: null  // Will be wrapped in a Proxy
    }
  },
  mounted() {
    this.map = new mapboxgl.Map({ ... })  // Proxy breaks GL internals
  }
}

// GOOD - Assign map as a plain instance property, not in data()
export default {
  mounted() {
    this.map = new mapboxgl.Map({
      container: this.$refs.mapContainer,
      center: [-71.05953, 42.3629],
      zoom: 13
    })
  },
  unmounted() {
    this.map?.remove()
  }
}

Why: In Vue (especially Vue 3), data() properties are wrapped in a Proxy for reactivity. Mapbox GL JS internally checks object identity and uses properties that don't survive proxy wrapping. Storing the map in data() causes subtle, hard-to-debug failures. Instead, assign the map instance directly as this.map in mounted() — properties assigned outside data() are not made reactive.

Mistake 5: Silent style/tile failures (no error handler)

javascript
// BAD — blank map when the token/style fails
const map = new mapboxgl.Map({ ... });

// GOOD — surface failures
map.on('error', (e) => {
  console.error(e.error || e);
  // optionally show an on-page error message
});

Mistake 6: Broken deck.gl CDN via jsDelivr +esm

html
<!-- BAD — often throws: does not provide export named 'makeBatchFromTable' -->
<script type="module">
  import { MapboxOverlay } from 'https://cdn.jsdelivr.net/npm/@deck.gl/mapbox@9.0.0/+esm';
</script>

<!-- GOOD — UMD bundle (or esm.sh) -->
<script src="https://unpkg.com/deck.gl@9.1.14/dist.min.js"></script>
<script>
  const { MapboxOverlay, ScatterplotLayer } = deck;
  map.addControl(
    new MapboxOverlay({
      interleaved: false,
      layers: [
        /* ... */
      ]
    })
  );
</script>

Use MapboxOverlay (Mapbox IControl), not a bare Deck as a map control.

Mistake 7: Draw toolbar without draw.create

If you load mapbox-gl-draw, listen for draw.create (and update the UI from draw.getAll()). Half-deleted handlers that leave a dangling }); crash the page.

Mistake 8: Layers lost after setStyle (no style.load rebind)

map.setStyle(...) replaces the style tree. Custom sources/layers/handlers added earlier are wiped unless you re-attach them.

javascript
function onStyleReady() {
  // re-add sources, layers, and interaction handlers here
}

map.on('style.load', onStyleReady);

document.querySelectorAll('[data-style]').forEach((btn) => {
  btn.addEventListener('click', () => {
    map.setStyle(btn.dataset.style);
    // do NOT only add layers on the first 'load' — wait for style.load after every switch
  });
});

Agent anti-pattern: style switcher buttons that call setStyle once with no style.load rebind. The first style works; every switch after looks broken.

Reference Files

Load these for framework-specific patterns and additional details:

  • references/vue.md — Vue Integration (mounted/unmounted lifecycle)
  • references/svelte.md — Svelte Integration (onMount/onDestroy)
  • references/angular.md — Angular Integration with SSR handling
  • references/vanilla.md — Vanilla JS (Vite) + Vanilla JS (CDN)
  • references/web-components.md — Web Components (basic + reactive + usage in React/Vue/Svelte)
  • references/nextjs.md — Next.js App Router + Pages Router
  • references/common-mistakes.md — Common Mistakes 4-7 + Testing Patterns
  • references/token-management.md — Token Management per bundler + Style Configuration

When to Use This Skill

Invoke this skill when:

  • Setting up Mapbox GL JS in a new project
  • Integrating Mapbox into a specific framework (React, Vue, Svelte, Angular, Next.js)
  • Building framework-agnostic Web Components
  • Creating reusable map components for component libraries
  • Debugging map initialization issues
  • Adding Mapbox Search functionality
  • Implementing proper cleanup and lifecycle management
  • Converting between frameworks (e.g., React to Vue)
  • Reviewing code for Mapbox integration best practices

Related Skills

  • mapbox-cartography: Map design principles and styling
  • mapbox-token-security: Token management and security
  • mapbox-style-patterns: Common map style patterns

Resources

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Mapbox Web Integration Patterns AI skill do?

Official integration patterns for Mapbox GL JS across popular web frameworks (React, Vue, Svelte, Angular). Covers setup, lifecycle management, token handling, search integration, and common pitfalls. Based on Mapbox's create-web-app scaffolding tool.

Why use Mapbox Web Integration Patterns on TypingMind?

Because you install it once and use it with any model. Mapbox Web Integration Patterns is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Mapbox Web Integration Patterns in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/mapbox/mapbox-agent-skills/tree/main/skills/mapbox-web-integration-patterns. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Mapbox Web Integration Patterns?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Mapbox Web Integration Patterns?

As many as you like. As long as a model supports skills, you can use Mapbox Web Integration Patterns with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Mapbox Web Integration Patterns AI skill free?

Yes. It is published on GitHub by mapbox under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇