slop.js

Never trust your own code again with slop.js, have AI generate your code each time you want to do something.

It will .eval() the code you prompt it to make every single fucking time. Oh there's also no sandboxing or protection.

Why?

I don't know.

When?

Hopefully never.

What (the fuck)?

Any and all further questions will go through my lawyer.

How it works

1. You give slop() a prompt.
2. It asks an AI model to write the JavaScript code.
3. It eval()s the generated code completely raw with no protection.
4. If the code throws, the error is fed back to the AI and it tries again, up to maxRetries times.
5. The function is async, so you can await it before running the next call.

Installation

npm install @gizzixz/slop.js

Usage

const slop = require('slop.js');

// Set your OpenAI API key (or pass it via the options object)
process.env.OPENAI_API_KEY = 'sk-...';

(async () => {
  // Generates code, evals it, returns the result.
  // If the code errors, the AI fixes it automatically.
  const result = await slop('write code that returns the sum of 1 through 10');
  console.log(result); // 55

  // You can await each call before moving on to the next.
  const greeting = await slop('write code that returns "Hello, world!"');
  console.log(greeting); // Hello, world!
})();

Options

const result = await slop('your prompt here', {
  provider:   'openai',   // AI provider: "openai" (default) or "anthropic"
  apiKey:     'sk-...',   // API key (default: OPENAI_API_KEY or ANTHROPIC_API_KEY env var)
  baseURL:    '...',      // Custom base URL for OpenAI-compatible providers (see below)
  model:      'gpt-4o',   // Model to use (default: "gpt-4o" for OpenAI, "claude-opus-4-5" for Anthropic)
  maxRetries: 10,         // Max AI fix attempts before throwing (default: 10)
});

Supported providers

OpenAI (default)

process.env.OPENAI_API_KEY = 'sk-...';
const result = await slop('your prompt');

Anthropic (Claude)

process.env.ANTHROPIC_API_KEY = 'sk-ant-...';
const result = await slop('your prompt', { provider: 'anthropic' });

OpenAI-compatible providers

Any provider with an OpenAI-compatible API (Groq, Mistral, Together AI, Perplexity, OpenRouter, Google Gemini, etc.) can be used via the baseURL option:

// Groq
const result = await slop('your prompt', {
  apiKey:   process.env.GROQ_API_KEY,
  baseURL:  'https://api.groq.com/openai/v1',
  model:    'llama-3.3-70b-versatile',
});

// Mistral
const result = await slop('your prompt', {
  apiKey:   process.env.MISTRAL_API_KEY,
  baseURL:  'https://api.mistral.ai/v1',
  model:    'mistral-large-latest',
});

// Google Gemini (OpenAI-compatible endpoint)
const result = await slop('your prompt', {
  apiKey:   process.env.GEMINI_API_KEY,
  baseURL:  'https://generativelanguage.googleapis.com/v1beta/openai/',
  model:    'gemini-2.0-flash',
});

⚠️ Warning

slop.js calls eval() on AI-generated code with zero sandboxing. It will execute whatever the model returns with full access to your Node.js process. Only use it in environments where you accept this risk.