Fetch web pages via HTTP GET and extract readable content in markdown, text, or raw modes.
npx skills add https://github.com/Acosmi/CrabClaw --skill web-fetch-acosmi
Please help me install this Agent Skill. Skill: web-fetch Source: https://github.com/Acosmi/CrabClaw/tree/main/docs/skills/tools/web/web-fetch Command: npx skills add https://github.com/Acosmi/CrabClaw --skill web-fetch-acosmi
Fetch and extract readable content from web pages by performing HTTP GET requests, stripping navigation, ads, and sidebars, and returning clean text in multiple formats with a lightweight 15-minute cache to speed repeated reads.
Provide a URL and optional extract_mode to fetch and extract readable content from the page.
High-intent search queries and answers about installing and using this skill.
To extract readable text from a web page URL, provide the link to fetch content via HTTP GET and strip navigation or ads to return clean markdown or text output.
Fetching web content and converting HTML to markdown is handled by performing an HTTP GET request with a markdown extraction mode to return readable output.
Yes, web content extraction caches results for 15 minutes to improve performance and speed up repeated reads of the same URL without making new HTTP GET requests.
Yes, you can get raw web content by specifying the raw extraction mode, bypassing the standard cleaning that strips navigation and ads from the HTML response.
Limitations include enforced URL validation and content-type checks, and the tool only supports HTTP GET requests to fetch URL content, returning a structured payload with status and length.
You need raw extraction mode when you want the original web content payload before cleaning, rather than the processed text or markdown output that strips sidebars and navigation.
AstorYHRetrieve data from a URL via HTTP GET with error handling.
CodingHeaderExtract web page content into markdown, JSON, or text outputs.
northseadlExtract content from modern websites and export clean Markdown documentation.
Wike-CHIFetch blocked webpage content as markdown via r.jina.ai, markdown.new, or defuddle.md.
