webMCP

@webmcp/core

JavaScript & TypeScript Library

Core library for parsing, validating, and optimizing web forms for AI processing. Reduce token usage by 67.6% with intelligent optimization algorithms while maintaining response quality and security.

25K+
Downloads
2.3K
GitHub Stars
95%+
Test Coverage
67.6%
Token Reduction

Quick Install

npm install @webmcp/core
View on GitHubDocumentation

Core Features

Everything you need to optimize web forms for AI processing

HTML Parser & Validator

Advanced HTML parsing with comprehensive webMCP schema validation

AI Optimization Engine

Reduce token usage by 67.6% with model-specific optimization algorithms

Security Layer

CSRF protection, JWT authentication, and encryption for sensitive data

Performance Caching

High-performance LRU/TTL caching system with >90% hit rate

Code Examples

Get started quickly with these practical examples

Basic Usage

import { WebMCPProcessor } from '@webmcp/core';

const processor = new WebMCPProcessor();
const webmcpData = processor.parseWebMCP(htmlContent);
const optimized = processor.optimizeForAI(webmcpData.elements, {
  target_model: 'gpt-4o',
  compression_level: 'advanced'
});

console.log(`Saved ${optimized.token_savings}% tokens`);

Advanced Configuration

import { WebMCPProcessor, OptimizationOptions } from '@webmcp/core';

const processor = new WebMCPProcessor({
  enableCaching: true,
  securityOptions: {
    enableCSRF: true,
    jwtSecret: process.env.JWT_SECRET
  }
});

const options: OptimizationOptions = {
  target_model: 'claude-3.5-sonnet',
  compression_level: 'premium',
  custom_rules: [
    { selector: 'input[type="password"]', priority: 'high' },
    { selector: '.optional', priority: 'low' }
  ]
};

const result = await processor.optimizeForAI(elements, options);

TypeScript Usage

import { 
  WebMCPProcessor, 
  WebMCPElement, 
  OptimizationResult 
} from '@webmcp/core';

const processor = new WebMCPProcessor();

interface FormData {
  elements: WebMCPElement[];
  goal: string;
}

async function optimizeForm(formData: FormData): Promise<OptimizationResult> {
  return await processor.optimizeForAI(formData.elements, {
    target_model: 'gpt-4o',
    goal: formData.goal,
    compression_level: 'advanced'
  });
}

API Reference

Complete API documentation for @webmcp/core

Main Classes

WebMCPProcessor

parseWebMCP(html)

Parse HTML and extract webMCP elements

optimizeForAI(elements, options)

Optimize elements for AI processing

validateSchema(data)

Validate against webMCP schema

calculateTokens(text, model)

Calculate token count for specific model

Configuration

OptimizationOptions

target_modelstring

AI model to optimize for (gpt-4o, claude-3.5-sonnet, etc.)

compression_levelstring

basic | advanced | premium

goalstring

Description of the form's purpose

custom_rulesarray

Custom optimization rules

Ready to Start Optimizing?

Install @webmcp/core today and reduce your AI token costs by 67.6%

Get StartedView Documentation