Download the PHP package maizied/roast-api without Composer

🥵 Roast as a Service (RaaS)

Random developer roasts, delivered via a blazing‑fast static API on GitHub Pages. Plug it into your apps, bots, terminals, or CI logs when “nice error messages” just aren’t enough.

Part of the Lorapok Ecosystem — Building the future of AI-driven developer tools. 🐛

🚀 What is RaaS?

Roast as a Service is a GitHub‑hosted, CDN‑backed JSON API that returns developer‑themed roasts in English and Bangla.
It’s completely static (no servers), but feels dynamic thanks to a smart client SDK that shards, caches, and randomly selects roasts on the fly.

Use it when you want your:

• CLI tools to roast the user on failure.
• CI pipeline to roast you when tests fail.
• Discord/Slack bots to respond with spicy dev insults.
• Portfolio or landing page to show random roasts on each visit.

🌐 Live Service

• Website: https://maijied.github.io/roast-as-a-service/
• API Base Path: https://maijied.github.io/roast-as-a-service/api/

⚡ Quick Start

1. Install via npm

2. Script Include (Browser)

Load the client SDK directly in your browser:

3. Install via Composer (PHP)

4. Install via pip (Python)

5. Direct Fetch

Or just fetch the JSON files directly:

🛠 How it works

RaaS exposes sharded JSON datasets over GitHub Pages, then a tiny client SDK picks, filters, and caches roasts in the browser, giving you an API‑like experience with pure static hosting.

• Static API: Roasts are stored in language‑specific shards (en, bn) and served as JSON over GitHub Pages’ global CDN for low TTFB.
• Smart client: The bundled client fetches a small shard, caches it, and returns random roasts with optional intensity and length filters.
• Zero ops: No servers, no cold starts, no scaling issues. Push to main, let Pages deploy and cache everything at the edge.

📦 API Overview

This is a static API: data comes from versioned JSON files served over CDN, and a lightweight JS client handles randomness, filtering and caching in the browser.

Endpoints (static JSON)

• Manifest:
  GET https://maijied.github.io/roast-as-a-service/api/manifest.json

• English Shard #1:
  GET https://maijied.github.io/roast-as-a-service/api/en/roasts-en-1.json

• Bangla Shard #1:
  GET https://maijied.github.io/roast-as-a-service/api/bn/roasts-bn-1.json

Example Response Structure

Each shard looks like:

Usage Example (curl + jq)

🧬 Why this architecture is “state of the art”

Roast as a Service is intentionally built as a static API on top of GitHub Pages and a CDN, instead of a traditional backend. For this use case (serving pre‑made content), this gives you the same “API experience” with less cost, less complexity, and better global latency.

Key points:

• Static JSON over CDN
  All responses are just versioned JSON files served by GitHub Pages’ global edge network, which is highly cache‑friendly and extremely fast for read‑heavy traffic.

• Zero backend, zero cold starts
  There is no application server to boot, scale, or warm up. Every request hits static content that can be served from edge cache with minimal TTFB.

• Sharded data layout
  Roasts are split into language‑specific shards (e.g. en/roasts-en-1.json) so each response stays small and cacheable even if the total dataset becomes large.

• Client‑side selection and filtering
  A tiny SDK handles randomness, intensity filters, and length limits in the client, so the “API” stays read‑only and ultra fast instead of computing on every request.

• Edge‑friendly cache behavior
  Because the content is static and not personalized, CDN nodes can cache responses aggressively without worrying about user‑specific data. That’s exactly what CDNs are optimized for.

• Version‑controlled API
  All JSON lives in Git; you can roll back, branch, and review changes like any other codebase while GitHub Pages redeploys and re‑caches automatically.

This combination (static JSON + sharding + client‑side logic + CDN caching) is effectively the “fast path” that many dynamic APIs end up approximating with layers of caching—here it’s the default.

⚡ Performance & Benchmarks

RaaS is built for extreme speed and low overhead. By sharding data and relying on CDN edge caching + local browser caching, it achieves massive throughput.

Latest Benchmark Results:

• Requests/sec: ~2,000 (Tested with 5,000 total requests)
• Success Rate: 100% (Zero failures under high concurrency)
• Avg Latency: ~58ms
• P95 Latency: ~159ms
• Cache Efficiency: 90%+ (Manifest and shard reuse)

[!TIP] This "Static API" approach beats the latency of traditional dynamic APIs by removing backend processing, cold starts, and database queries from the request path.

🐛 More from Lorapok

RaaS is a proud member of the Lorapok family. Check out our other tools:

• Lorapok CLI: The intelligent AI agent framework for developers.
• Lorapok Media Player: A modern, feature-rich media player built for power users.

👤 Author

Maizied
📧 [email protected]
🔗 GitHub · npm

📄 License

MIT © Maizied