parallel-web-extract
Extrai conteúdo de múltiplas URLs em paralelo, de forma eficiente em tokens. Lida com páginas web, artigos, PDFs e sites com muito JavaScript em um único comando. Executa em um contexto bifurcado para minimizar a sobrecarga de tokens em comparação com o WebFetch integrado. Suporta extração em lote de várias URLs com objetivos de foco opcionais. Requer instalação e autenticação do parallel-cli; gera o conteúdo extraído como markdown em um arquivo local para consultas de acompanhamento.
npx skills add https://github.com/parallel-web/parallel-agent-skills --skill parallel-web-extractURL Extraction
Extract content from: $ARGUMENTS
Command
Choose a short, descriptive filename based on the URL or content (e.g., vespa-docs, react-hooks-api). Use lowercase with hyphens, no spaces. Substitute it into the command inline — $FILENAME is a placeholder, not a shell variable.
parallel-cli extract "$ARGUMENTS" --json -o "/tmp/$FILENAME.json"
Concrete example:
parallel-cli extract "https://docs.parallel.ai" --json -o "/tmp/parallel-docs.json"
Note: -o always saves JSON. The extension must be .json.
Options if needed:
--objective "focus area"to focus extraction on a specific goal (also silences the "neither objective nor search_queries" warning that V1 emits when neither is set)-q "keyword"(repeatable) to prioritize keywords in excerpts--full-contentto include the complete page body (for long articles, PDFs, or when excerpts may not capture what you need)--full-content-max-chars Nto cap full-content size per result--no-excerptsto strip excerpts when you only want full content
Handling failed extractions
If the response has an errors field, an empty results array, or a 404/timeout for the URL, do NOT fabricate content. Tell the user the extraction failed, surface the upstream status, and suggest:
- Verifying the URL (the page may have moved)
- Retrying with
--full-contentif excerpts came back empty but the page exists - Using
parallel-cli searchto locate the current URL if the page was renamed
Response format
Return content as:
Then the extracted content verbatim, with these rules:
- Keep content verbatim - do not paraphrase or summarize
- Parse lists exhaustively - extract EVERY numbered/bulleted item
- Strip only obvious noise: nav menus, footers, ads
- Preserve all facts, names, numbers, dates, quotes
After the response, mention the output file path (/tmp/$FILENAME.json) so the user knows it's available for follow-up questions.
Setup
If parallel-cli is not found, install and authenticate:
/parallel:parallel-cli-setup
If parallel-cli extract returns 403, tell the user balance is likely required. Offer to run parallel-cli balance get, and if needed ask for explicit confirmation before running parallel-cli balance add <amount_cents>. Then retry the original extract command.