Bun Hot Reloading logo

Bun Hot Reloading

Community
secondsky
bun-hot-reloading

Use when implementing hot reloading with Bun (--hot, --watch), HMR, or automatic code reloading during development. Covers watch mode, hot mode, and HTTP server reload.

Overview

Publishersecondsky
Repositoryclaude-skills
Skill namebun-hot-reloading
Stars
219
Forks
31
Bundled files
Instructions only
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

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

Installation

Install the Bun Hot Reloading 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/secondsky/claude-skills.git /tmp/claude-skills
mkdir -p .claude/skills
cp -r /tmp/claude-skills/plugins/bun/skills/bun-hot-reloading .claude/skills/bun-hot-reloading
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Bun Hot Reloading 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 Bun Hot Reloading 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 Bun Hot Reloading 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.

Bun Hot Reloading

Bun provides built-in hot reloading for faster development cycles.

Watch Mode vs Hot Mode

Feature--watch--hot
BehaviorRestart processReload modules
StateLost on reloadPreserved
Speed~20ms restartInstant reload
Use caseAny file typeBun.serve HTTP

Watch Mode (--watch)

Restarts the entire process when files change.

bash
# Basic watch mode
bun --watch run src/index.ts

# Watch specific script
bun --watch run dev

# Watch with test runner
bun --watch test

package.json Scripts

json
{
  "scripts": {
    "dev": "bun --watch run src/index.ts",
    "dev:server": "bun --watch run src/server.ts",
    "test:watch": "bun --watch test"
  }
}

Watch Behavior

  • Watches imported files automatically
  • Triggers on any .ts, .tsx, .js, .jsx change
  • Also watches .json imports
  • Restarts with fresh state

Hot Mode (--hot)

Reloads modules in-place without restarting the process.

bash
bun --hot run src/server.ts

HTTP Server Hot Reload

typescript
// src/server.ts
let counter = 0; // State preserved across hot reloads

export default {
  port: 3000,
  fetch(req: Request) {
    counter++;
    return new Response(`Request #${counter}`);
  },
};
bash
bun --hot run src/server.ts

When you modify server.ts, the module reloads instantly while counter keeps its value.

Bun.serve with Hot Reload

typescript
// src/server.ts
const server = Bun.serve({
  port: 3000,
  fetch(req) {
    return new Response("Hello!");
  },
});

// Hot reload handler
if (import.meta.hot) {
  import.meta.hot.accept(() => {
    console.log("Hot reload!");
  });
}

console.log(`Server running on port ${server.port}`);

import.meta.hot API

typescript
// Check if hot reload is available
if (import.meta.hot) {
  // Accept updates to this module
  import.meta.hot.accept();

  // Accept with callback
  import.meta.hot.accept((newModule) => {
    console.log("Module updated:", newModule);
  });

  // Cleanup before reload
  import.meta.hot.dispose(() => {
    // Close connections, clear intervals, etc.
    clearInterval(myInterval);
  });

  // Decline hot reload (force full restart)
  import.meta.hot.decline();

  // Invalidate this module (trigger parent reload)
  import.meta.hot.invalidate();
}

HTTP Server Patterns

Express-like Pattern

typescript
// src/server.ts
import { createApp } from "./app";

const app = createApp();

const server = Bun.serve({
  port: 3000,
  fetch: app.fetch,
});

// Hot reload: recreate app
if (import.meta.hot) {
  import.meta.hot.accept((newModule) => {
    // Reload with new fetch handler
    server.reload({
      fetch: newModule.default.fetch,
    });
  });
}

Stateful Server

typescript
// src/server.ts
// Store in globalThis to survive reloads
globalThis.connections ??= new Set();

const server = Bun.serve({
  port: 3000,
  fetch(req) {
    return new Response(`Connections: ${globalThis.connections.size}`);
  },
  websocket: {
    open(ws) {
      globalThis.connections.add(ws);
    },
    close(ws) {
      globalThis.connections.delete(ws);
    },
  },
});

if (import.meta.hot) {
  import.meta.hot.accept();
}

Custom Watch Implementation

typescript
// dev-server.ts
import { watch } from "fs";

const srcDir = "./src";
let server: ReturnType<typeof Bun.serve> | null = null;

async function startServer() {
  // Dynamic import with cache busting
  const module = await import(`./src/server.ts?t=${Date.now()}`);

  if (server) {
    // Bun 1.2+: server.stop() returns a Promise — await it so the old
    // socket is fully torn down before we rebind the same port.
    await server.stop();
  }

  server = Bun.serve(module.default);
  console.log(`Server started on port ${server.port}`);
}

// Initial start
await startServer();

// Watch for changes
watch(srcDir, { recursive: true }, async (event, filename) => {
  if (filename?.endsWith(".ts") || filename?.endsWith(".tsx")) {
    console.log(`\n[${event}] ${filename}`);
    await startServer();
  }
});

console.log("Watching for changes...");

WebSocket Live Reload

Server

typescript
// src/dev-server.ts
const clients = new Set<ServerWebSocket>();

const server = Bun.serve({
  port: 3000,
  fetch(req, server) {
    if (req.headers.get("upgrade") === "websocket") {
      server.upgrade(req);
      return;
    }

    // Inject reload script in dev
    const html = `
      <!DOCTYPE html>
      <html>
        <body>
          <h1>Hello!</h1>
          <script>
            const ws = new WebSocket('ws://localhost:3000');
            ws.onmessage = (e) => {
              if (e.data === 'reload') location.reload();
            };
          </script>
        </body>
      </html>
    `;
    return new Response(html, {
      headers: { "Content-Type": "text/html" },
    });
  },
  websocket: {
    open(ws) {
      clients.add(ws);
    },
    close(ws) {
      clients.delete(ws);
    },
  },
});

// Notify clients on file change
watch("./src", { recursive: true }, () => {
  clients.forEach((ws) => ws.send("reload"));
});

Vite Integration

For frontend development with HMR:

typescript
// vite.config.ts
import { defineConfig } from "vite";
import react from "@vitejs/plugin-react";

export default defineConfig({
  plugins: [react()],
  server: {
    port: 5173,
    hmr: true,
  },
});
bash
# Use Bun to run Vite
bunx --bun vite

Testing with Watch

bash
# Watch tests
bun --watch test

# Watch specific file
bun --watch test src/utils.test.ts

# With bail (stop on first failure)
bun --watch test --bail

Environment Detection

typescript
// Check if running with --hot
const isHot = !!import.meta.hot;

// Check if running with --watch
const isWatch = process.env.BUN_WATCH === "1";

// Development mode
const isDev = process.env.NODE_ENV !== "production";

if (isDev) {
  console.log("Running in development mode");
  console.log(`Hot reload: ${isHot}`);
  console.log(`Watch mode: ${isWatch}`);
}

Common Issues

State Not Preserved

typescript
// ❌ State lost on hot reload
let cache = new Map();

// ✅ State preserved on hot reload
globalThis.cache ??= new Map();
const cache = globalThis.cache;

Cleanup Not Running

typescript
// ❌ Interval keeps running after reload
setInterval(() => console.log("tick"), 1000);

// ✅ Clean up on dispose
const interval = setInterval(() => console.log("tick"), 1000);

if (import.meta.hot) {
  import.meta.hot.dispose(() => {
    clearInterval(interval);
  });
}

Module Not Reloading

typescript
// ❌ Import not watched
const config = require("./config.json");

// ✅ Use import for watching
import config from "./config.json";

Common Errors

ErrorCauseFix
Changes not detectedFile not importedCheck import chain
State lostUsing --watchUse --hot or globalThis
Port in useServer not stoppedawait server.stop() before rebind
Memory leakNo cleanupUse dispose callback

When to Load References

Load references/advanced-hmr.md when:

  • Custom HMR protocols
  • Module federation
  • Complex state management

Load references/debugging.md when:

  • HMR not working
  • State issues
  • Performance debugging

Frequently asked questions

What does the Bun Hot Reloading AI skill do?

Use when implementing hot reloading with Bun (--hot, --watch), HMR, or automatic code reloading during development. Covers watch mode, hot mode, and HTTP server reload.

Why use Bun Hot Reloading on TypingMind?

Because you install it once and use it with any model. Bun Hot Reloading 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 Bun Hot Reloading in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/secondsky/claude-skills/tree/main/plugins/bun/skills/bun-hot-reloading. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Bun Hot Reloading?

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 Bun Hot Reloading?

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

Is the Bun Hot Reloading AI skill free?

Yes. It is published on GitHub by secondsky 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 👇