adscrawl-browser
Extract and Automate Websites with AdsCrawl
Modern websites often require browser rendering before useful content appears. This skill uses AdsCrawl for extraction, screenshots, and stateful browser sessions.
自分のエージェントでインストール
このリクエストをエージェントにコピーしてください。正規の Skill ページとマニフェストが含まれています。
Review the Skillstore skill "adscrawl-browser" from https://skillstore.io/skills/adscrawl-adscrawl-browser.md and its manifest at https://skillstore.io/api/skills/adscrawl-adscrawl-browser/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.エージェントは引き続き計画を提示し、セキュリティポリシーで必要な確認を求める必要があります。
エージェントが読めるリソース
AI エージェント、クローラー、スクリプトがページ全体ではなく整理されたコンテキストを必要とする場合は、これらのリンクを使ってください。
テストする
「adscrawl-browser」を使用しています。 Extract a rendered product article as Markdown and summarize it.
期待される結果:
Saved the rendered article as page.md. The summary covers the product purpose, primary features, pricing details, and stated availability.
「adscrawl-browser」を使用しています。 Capture a full-page screenshot for visual review.
期待される結果:
- Saved a non-empty 1440 by 900 full-page PNG.
- Verified that the page loaded without a challenge screen.
- Noted one subscription overlay covering part of the article.
「adscrawl-browser」を使用しています。 Use a remote browser to complete an authorized multi-step form check.
期待される結果:
Completed the requested navigation and validation steps in one CDP session. Closed the remote session after recording the observed results.
セキュリティ監査
安全All 45 static findings are false positives caused by Markdown backticks, documented environment variables, and the declared AdsCrawl API endpoint. The reviewed skill contains no embedded credentials, command injection, prompt injection, or unauthorized data handling.
リスク要因
⚙️ 外部コマンド (36)
🌐 ネットワークアクセス (2)
このレポートを共有・引用
バージョン付き評価レポート、中立的なバッジ、埋め込みカード、引用を共有できます。Skillstore は証拠を報告しますが、この Skill が安全かどうかは判断しません。
レポートリンクをコピー
https://skillstore.io/skills/adscrawl-adscrawl-browser/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdownバッジ
[](https://skillstore.io/skills/adscrawl-adscrawl-browser?utm_source=security_passport_badge)HTMLバッジ
<a href="https://skillstore.io/skills/adscrawl-adscrawl-browser?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/adscrawl-adscrawl-browser/security.svg" alt="Skillstore security assessment" loading="lazy"></a>埋め込みカード
<iframe src="https://skillstore.io/embed/skills/adscrawl-adscrawl-browser.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>学術引用 (APA · BibTeX · CFF)
APA形式の引用
adscrawl. (2026). adscrawl-browser security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/adscrawl-adscrawl-browser/audits/1BibTeX形式の引用
@techreport{adscrawl-adscrawl-adscrawl-browser-2026,
author = {adscrawl},
title = {adscrawl-browser security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/adscrawl-adscrawl-browser/audits/1},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "adscrawl-browser security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
- name: "adscrawl"
date-released: "2026-08-26"
url: "https://skillstore.io/skills/adscrawl-adscrawl-browser/audits/1"
identifiers:
- type: other
value: "skillstore:adscrawl-adscrawl-browser:audit:1"
description: "Skillstore immutable audit report identifier"
Skillstore スコア
このスコアの理由 証拠の信頼度: 中作成できるもの
Research rendered web content
Extract clean Markdown from JavaScript-heavy pages for analysis, summaries, or retrieval workflows.
Verify website appearance
Capture full-page or viewport screenshots with requested locale, timezone, and viewport settings.
Automate stateful browser tasks
Create a remote CDP session for authorized multi-step navigation, forms, login flows, or debugging.
これらのプロンプトを試す
Extract TARGET_URL as Markdown. Save the result, verify it contains meaningful page content, and summarize the main points.
Capture a full-page PNG of TARGET_URL at 1440 by 900. Check the image for overlays, missing assets, or incomplete loading.
Extract TARGET_URL for COUNTRY_CODE and capture a matching screenshot. Compare visible content, pricing, and availability without using a custom proxy.
Create one CDP session for TARGET_URL, complete AUTHORIZED_STEPS with the project's browser library, record outcomes, and always close the session.
ベストプラクティス
- Use extraction or screenshots when interaction is unnecessary, and reserve CDP sessions for stateful tasks.
- Keep credentials in environment variables and exclude token-bearing URLs, cookies, and proxy secrets from logs.
- Verify saved outputs and close every CDP session in a cleanup path.
回避
- Do not combine managed country routing with a custom proxy in one request.
- Do not retry repeatedly when the page, credentials, or service returns a persistent error.
- Do not access authenticated content, personal data, or external side effects without explicit authorization.