SkillProof tested Claude skills

Browser Search

Stealth browser fetch a multi-engine search pro AI web research

od Johell1NS · Johell1NS/browser-search

Funguje s nastavením ★ 8.4/10

Browser Search — Stealth browser fetch a multi-engine search pro AI web research

Co umí

Třívrstvý web research stack: SearXNG pro multi-engine search, Camofox pro JS-heavy pages a CloakBrowser pro sites behind Cloudflare nebo DataDome. Spouští se na web research, scraping a pages, které plain fetching nemůže renderovat.

Testovací report

Úkol: extract title a text z JS-rendered page quotes.toscrape.com/js — baseline s bare curl proti cloak-fetch.mjs bundle. Baseline byl useless: po tag removal, 0 quotes (content byl client-side rendered); skill launched stealth-Chromium a vrátil title 'Quotes to Scrape', 1499 characters, a all 10 quotes za 6.6 s. Security claims byly verified live, not by text: SSRF-block fungoval na 169.254.169.254 ('Blocked IP') a localhost ('Blocked hostname'), path-traversal guard odrazil --script /etc/passwd, a bundled tests daly 28/28 a 12/12 passed. Main caveat: 2 out of 3 tools (SearXNG a Camofox) required Docker, který nebyl v environmentu — I DID NOT measure them; flagship 'multi-engine search' zůstal unchecked, a jeho own check.sh honestly showed this. Another inaccuracy: SKILL.md sliboval 'all commands/scripts are read-only, safe in Plan mode', ale very first run of cloak-fetch downloaded 140 MB of Chromium do ~/.cloakbrowser. Docker + two containers — real manual work, hence setup.

Testováno: 2026-07-17 · Claude Code 2.x (agent harness)

Instalace

git clone https://github.com/Johell1NS/browser-search.git
mkdir -p ~/.claude/skills
cd browser-search && rm -rf .git && cp -r . ~/.claude/skills/browser-search
cd ~/.claude/skills/browser-search && npm install

Příkazy a ukázkové prompty

  • /browser-searchStealth browser fetch a multi-engine search pro AI web research

Skilly se spouštějí běžnými požadavky — žádné příkazy k zapamatování. Po instalaci ho aktivují prompty jako tyto (anglicky):

  • This page needs JavaScript to render, can you fetch the actual content
  • Search across multiple engines for me and pull the real page text back
  • This site is behind Cloudflare, I still need the rendered page content

Kompletní testovací report (anglicky) →