#!/usr/bin/env npx tsx
/**
 * MCP Web Scraper Tool — Monetized with SettleGrid
 *
 * A complete MCP server that scrapes and extracts data via Firecrawl.
 * Fork this template, add your Firecrawl API key, and deploy.
 *
 * Setup:
 *   1. npm install @settlegrid/mcp
 *   2. Set FIRECRAWL_API_KEY and SETTLEGRID_API_KEY in your env
 *   3. Register your tool at settlegrid.ai/dashboard/tools
 *   4. Run: npx tsx mcp-web-scraper.ts
 *
 * Pricing: 3 cents per scrape, 5 cents per structured extraction, 2 cents per batch URL
 *   - Firecrawl costs ~$0.004/page on Starter plan (500 credits at $19)
 *   - 3 cents gives you ~7.5x margin on simple scrape
 *   - Structured extraction uses more compute, 5 cents = ~5x margin
 *   - Batch discount at 2 cents/URL encourages higher volume
 *
 * Revenue: You keep 95-100% (100% on Free tier, 95% on paid tiers)
 */

import { settlegrid } from '@settlegrid/mcp'

// ── SettleGrid Setup ────────────────────────────────────────────────────────

const sg = settlegrid.init({
  toolSlug: 'my-web-scraper', // Replace with your tool slug
  pricing: {
    defaultCostCents: 3,
    methods: {
      scrape_url: { costCents: 3, displayName: 'Scrape URL' },
      extract_structured: { costCents: 5, displayName: 'Structured Extraction' },
      scrape_batch: { costCents: 2, displayName: 'Batch Scrape (per URL)' },
    },
  },
})

// ── Firecrawl API Helper ────────────────────────────────────────────────────

const FIRECRAWL_BASE = 'https://api.firecrawl.dev/v1'
const URL_RE = /^https?:\/\/[^\s/$.?#].[^\s]*$/

async function firecrawlPost(path: string, body: Record<string, unknown>): Promise<Record<string, unknown>> {
  const response = await fetch(`${FIRECRAWL_BASE}${path}`, {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json',
      Authorization: `Bearer ${process.env.FIRECRAWL_API_KEY!}`,
    },
    body: JSON.stringify(body),
  })
  if (!response.ok) {
    throw new Error(`Firecrawl API returned ${response.status}: ${response.statusText}`)
  }
  return response.json()
}

function validateUrl(url: string): void {
  if (!url || !URL_RE.test(url)) throw new Error('A valid HTTP or HTTPS URL is required')
}

// ── Scraper Methods ─────────────────────────────────────────────────────────

interface ScrapeArgs { url: string; includeLinks?: boolean }

async function scrapeUrl(args: ScrapeArgs): Promise<{
  result: { markdown: string; title: string; url: string; wordCount: number }
}> {
  validateUrl(args.url)
  const data = await firecrawlPost('/scrape', {
    url: args.url, formats: ['markdown'], onlyMainContent: true, includeLinks: args.includeLinks ?? false,
  })
  const doc = data.data as Record<string, unknown>
  const markdown = (doc?.markdown as string) ?? ''
  return {
    result: {
      markdown, title: (doc?.metadata as Record<string, string>)?.title ?? '',
      url: args.url, wordCount: markdown.split(/\s+/).filter(Boolean).length,
    },
  }
}

interface ExtractArgs { url: string; schema: Record<string, string> }

async function extractStructured(args: ExtractArgs): Promise<{ data: Record<string, unknown> }> {
  validateUrl(args.url)
  if (!args.schema || Object.keys(args.schema).length === 0) {
    throw new Error('An extraction schema with at least one field is required')
  }
  const data = await firecrawlPost('/scrape', {
    url: args.url, formats: ['extract'], extract: { schema: args.schema },
  })
  const doc = data.data as Record<string, unknown>
  return { data: (doc?.extract as Record<string, unknown>) ?? {} }
}

interface BatchArgs { urls: string[] }

async function scrapeBatch(args: BatchArgs): Promise<{
  batch: { results: Array<{ url: string; markdown: string; title: string }>; successCount: number; failCount: number }
}> {
  if (!args.urls || args.urls.length === 0) throw new Error('At least one URL is required')
  if (args.urls.length > 10) throw new Error('Batch size is limited to 10 URLs per call')
  for (const url of args.urls) validateUrl(url)

  const results: Array<{ url: string; markdown: string; title: string }> = []
  let failCount = 0
  for (const url of args.urls) {
    try {
      const data = await firecrawlPost('/scrape', { url, formats: ['markdown'], onlyMainContent: true })
      const doc = data.data as Record<string, unknown>
      results.push({ url, markdown: (doc?.markdown as string) ?? '', title: (doc?.metadata as Record<string, string>)?.title ?? '' })
    } catch { failCount++ }
  }
  return { batch: { results, successCount: results.length, failCount } }
}

// ── Wrap with SettleGrid Billing ─────────────────────────────────────────────

export const billedScrape = sg.wrap(scrapeUrl, { method: 'scrape_url' })
export const billedExtract = sg.wrap(extractStructured, { method: 'extract_structured' })
export const billedBatch = sg.wrap(scrapeBatch, { method: 'scrape_batch' })

// ── REST Alternative ────────────────────────────────────────────────────────
// import { settlegridMiddleware } from '@settlegrid/mcp/rest'
//
// const withBilling = settlegridMiddleware({
//   toolSlug: 'my-web-scraper',
//   pricing: {
//     defaultCostCents: 3,
//     methods: {
//       scrape_url: { costCents: 3 },
//       extract_structured: { costCents: 5 },
//       scrape_batch: { costCents: 2 },
//     },
//   },
// })
//
// export async function POST(request: Request) {
//   return withBilling(request, async () => {
//     const { url } = await request.json()
//     const result = await scrapeUrl({ url })
//     return Response.json(result)
//   }, 'scrape_url')
// }