@mayurajs/browser-firecrawl

Firecrawl for Mayura, two ways, from any runtime with no dependency:

• firecrawlBrowsers — browsers for mayura/browser, as Firecrawl Interact sessions driven over CDP.
• firecrawlTools — Firecrawl's web data as tools: scrape a page, map a site, and, when enabled, search the web, crawl a site or extract structured data.

npm install mayura @mayurajs/browser-firecrawl

import { firecrawlTools } from '@mayurajs/browser-firecrawl';

const tools = firecrawlTools({ apiKey, origins: ['https://docs.example.com'], crawl: true });
// Tools: web.scrape, web.map (web:web:read) and web.crawl (web:web:crawl).

Tools

| Tool | Permission | Option | | --- | --- | --- | | <name>.scrape: a page as markdown (the default), html, its links or a summary | web:<name>:read | always | | <name>.map: a site's pages, optionally matching a search | web:<name>:read | always | | <name>.search: web search results (titles, URLs, descriptions) | web:<name>:search | search: true | | <name>.crawl: up to maxCrawlPages pages of a site, as markdown | web:<name>:crawl | crawl: true | | <name>.extract: structured data from a page, by prompt and optional JSON Schema | web:<name>:extract | extract: true |

• origins is required: the sites the tools may read, as createBrowsers takes them, or 'all'. A URL outside them is refused, and comes back as the tool's result so the model can choose another. Search results are not bound by it, but reading them is. A crawl stays on its site, and pages a redirect took elsewhere are left out.
• Firecrawl fetches the pages from its own servers, with TLS checked (Firecrawl's own default skips it). Page content is the tools' result for the model to read: treat it as untrusted text.
• Content is cut at maxPageBytes (64 KiB). A crawl polls Firecrawl until done, and is cancelled when the call is, or after crawlTimeoutMs (5 minutes), so it stops spending. costMicros charges each tool's calls to the run's budget; Firecrawl bills credits on its side.
• Extraction uses Firecrawl's JSON format on one page, with its model.
• Options: apiKey, name (web), origins, search, crawl, extract, maxPageBytes, maxCrawlPages (10, at most 100), crawlTimeoutMs, costMicros, fetch, baseUrl.

Browsers

import { createBrowsers } from 'mayura/browser';
import { firecrawlBrowsers } from '@mayurajs/browser-firecrawl';

const browsers = createBrowsers(firecrawlBrowsers({ apiKey }), { maxBrowsers: 2, maxLifetimeMs: 10 * 60_000, origins: ['https://example.com'] });

• Each browser is an Interact session: created when opened, deleted when released. Firecrawl ends it at its lifetime (30 s to an hour) or after activityTimeoutSeconds (300) without activity.
• liveViewUrl is Firecrawl's live view: view-only by default; liveView: 'interact' gives the interactive one, which lets whoever has the link use the browser; false gives none.
• Treat the CDP URL as a secret, as Firecrawl's docs differ on what it carries: Mayura never logs it.
• Origins are enforced inside the browser, by intercepting its requests over CDP; a live check against Firecrawl confirmed it on October 6, 2026. Should a connection ever not allow that, opening a browser fails rather than running unchecked.
• Options: apiKey, liveView, activityTimeoutSeconds, maxLifetimeMs (1 hour), fetch, baseUrl.

See the browser guide. Apache-2.0.