Pular para o conteúdo
TakoAPI
Voltar às skills

browser-ladder

por ktpriyatham

Browser & Automation
2.2kDownloads informados pelo clawskills.sh
3Estrelas
0Visualizações
0 curtidaEntre para curtir esta skill
—Ainda sem avaliações
Sua avaliação

Resumo

Climb the browser ladder — start free, escalate.

Descrição

O que esta skill faz

Automatically selects the cheapest browser access method that works for a given URL. Starts with plain HTTP fetch, escalates through local Playwright, cloud-free BrowserCat, and finally Browserless.io for CAPTCHA bypass, only when lower rungs fail.

Instead of always paying for cloud browsers or always fighting with local Docker, it routes each request to the cheapest tool that can handle it.

Quando usá-la

  • Scraping a React SPA that needs JavaScript rendering
  • Bypassing Cloudflare bot detection on a target site
  • Extracting data from a static API endpoint without spinning up a browser
  • Falling back to cloud browser when Docker is unavailable locally
  • Accessing OAuth-protected or MFA-gated pages programmatically

Exemplo de fluxo de trabalho

ENTRADA

User asks: 'Scrape the pricing table from a Cloudflare-protected SPA'

AGENTE
  1. 1

    Attempts Rung 1 web_fetch — fails due to JavaScript rendering requirement

  2. 2

    Runs Rung 2 Playwright via Docker — blocked by Cloudflare bot detection

  3. 3

    Escalates to Rung 4 Browserless.io using BROWSERLESS_TOKEN

  4. 4

    Connects over CDP, CAPTCHA is handled automatically

  5. 5

    Returns rendered page content

SAÍDA

Full HTML of the pricing table extracted from the protected SPA

Requisitos

  • BROWSERCAT_API_KEY environment variable (free tier at browsercat.com, optional)
  • BROWSERLESS_TOKEN environment variable ($10/mo at browserless.io, optional)
  • Docker installed (for Rung 2 local Playwright)

Instalação

clawhub install ktpriyatham/browser-ladder

Execute o comando da CLI ou cole o prompt no Claude Code, Cursor, Codex ou no agente de código que você usa.

Links