メインコンテンツへスキップ

Get the details you need from web pages

Turn web pages into readable text and structured research material.

MCP / oo CLI + Skill3 件のサービス

できること

Read a page or a bounded set of pages, then extract the details relevant to your task. Useful outputs include product specifications, document outlines, and comparison tables with source URLs.

Collect page content

Keep the relevant headings and passages from a known webpage.

Extract a comparison table

Organize selected fields across pages while retaining a source URL for each row.

Find pages before extracting content

Map a site with search guidance and path filters, then extract the selected URLs.

Extract evidence for a specific question

Read selected URLs with an explicit research objective and retain the relevant passages.

Read a web page

Extract page text with Jina Reader.

プロンプトを試す

角括弧内の値を置き換え、接続済みの AI アシスタントにタスクをコピーしてください。

接続方法を選ぶ

Collect page content

Keep the relevant headings and passages from a known webpage.

Use OOMOL MCP with AIsa to complete this task: Use AIsa to read [HTTPS page URL]. Extract the sections about [subject] and organize them with headings and source links.

Extract a comparison table

Organize selected fields across pages while retaining a source URL for each row.

Use OOMOL MCP with Firecrawl to complete this task: Use my Firecrawl connection to read [page URLs]. Build a table of [fields], showing the source URL for each row.

Find pages before extracting content

Map a site with search guidance and path filters, then extract the selected URLs.

Use OOMOL MCP with AIsa to complete this task: Map [website] for pages about [topic] with a limit of [count]. Extract the relevant URLs in batches of up to 20 and retain any partial failures.

Extract evidence for a specific question

Read selected URLs with an explicit research objective and retain the relevant passages.

Use OOMOL MCP with AIsa to complete this task: Read [URLs] to answer [specific question]. Use Parallel extraction to collect relevant passages and preserve the source for each finding.

Extract page text with Jina Reader.

Use OOMOL MCP with OOMOL Fusion API to complete this task: Use OOMOL Jina Reader to read [page URL] and return the main text.

Web Extraction を使い始める

  1. AI アシスタントを接続する

    接続方法を選ぶ

    MCP クライアントに OOMOL を追加し、チームでサインインしてサービスを使い始めます。

  2. サービスを選ぶ

    AIsaChoose OOMOL-provided access when available, or connect your AIsa account.

    Keep the relevant headings and passages from a known webpage.

    AIsa の接続方法 →

    FirecrawlConnect your Firecrawl account to use the resources available to it.

    Organize selected fields across pages while retaining a source URL for each row.

    Firecrawl の接続方法 →

    OOMOL Fusion APISign in to OOMOL; no separate provider credential is required.

    Use OOMOL services with your current team.

    OOMOL Fusion API の接続方法 →
  3. 最初のタスクを試す

    Use OOMOL MCP with AIsa to complete this task: Use AIsa for this check: Read [HTTPS page URL] and show its title and the first relevant paragraph about [subject].

Web Extraction に対応するサービス

タスクに合うサービスを選びます。ツール数は各サービスの全ツールの合計です。

すべての操作

7 ツール

Scrape Firecrawl

Retrieve one HTTPS web page as Markdown through Firecrawl. The AIsa route supports basic proxy mode and excludes PDF URLs.

約 0.3213 クレジット / リクエスト

aisa.scrape_firecrawl

Map Firecrawl

Discover website links through Firecrawl. Set limit to bound discovery and usage before retrieving individual pages.

約 0.3213 クレジット / リクエスト

aisa.map_firecrawl

Scrape

Scrape a URL with Firecrawl or execute an enabled Alexandria provider tool and return the requested output formats.

firecrawl.scrape

Map Tavily

Discover site URLs using path and domain filters, crawl depth, and natural-language instructions.

約 19.2 クレジット / リクエスト

aisa.map_tavily

Extract Tavily

Extract content from known URLs through Tavily. Inspect failed_results for per-URL failures; use crawl_tavily to discover and retrieve pages from a website.

約 7.2 クレジット / リクエスト

aisa.extract_tavily

Extract Parallel

Extract relevant excerpts or full content from up to 20 URLs, guided by an objective and search terms. Return partial results and per-URL errors.

約 6 クレジット / リクエスト

aisa.extract_parallel

Jina Reader Read

Read and extract URL content

10 クレジット / 1,000,000 文字

fusion-api.jina_reader_read

よくある質問

When should I use extraction instead of search?

Use search to discover sources. Use extraction when you already have pages and need their text or specific details for a document or comparison.