Skip to content

usestring/string-web-access

v1.0.0MIT

Skills and the hosted MCP server for String Web Access: fetch pages, search the web, run site integrations, crawl sitemaps and send POST/PUT/PATCH requests.

string-fetch

Fetch any web page and get clean, LLM-ready Markdown back, via String Web Access. Use when the user gives you a URL, asks you to read, open, summarize or extract from a page, or when a search result needs its full content. Also supports custom headers, country-specific proxies and driving a real browser through clicks and scrolls before capture. Reads only — to send a POST, PUT or PATCH, use string-request.

string-request

Send a POST, PUT or PATCH with a body to a URL, through String Web Access. Use when the user wants to submit something to an endpoint rather than read a page — calling a JSON API that needs a payload, posting to a webhook, updating a record. For reading a page, use string-fetch instead.

string-search

Search the web with String Web Access and get structured Google results back. Use when the user wants to search the web, find articles or sources, look up current information, check recent news, or says "search for", "find me", "look up", "what's the latest on", "who is", or asks anything needing information from the live internet rather than training data. Returns organic results with position, title, URL, snippet and display URL.

string-sitemap

Discover every URL on a website with String Web Access. Use when the user wants a site's full URL list, asks to crawl or map a site, needs to find all pages of a given kind (every blog post, every product, every docs page), or when a task needs coverage of a whole site rather than one page. Runs as a quoted, approved, asynchronous job — quote first, approve the cost, then poll and page through results.

string-web-access

Best practices for getting web content with String Web Access. Load this before any multi-step web task — research, competitive analysis, monitoring a set of pages, or anything where one fetch is not enough. Covers the search → fetch → browser escalation rule, when each tool is the right one, how to handle blocked or empty responses, and how to keep cost and latency down. Also load when a fetch returns a block page, a CAPTCHA, or an empty body and you need to know what to change.