markdown-fetch
от howtimeschange
Web & Frontend DevelopmentКратко
Optimizes web fetching by using Cloudflare's Markdown for Agents, reducing token consumption by ~80%.
Описание
Что делает этот навык
Wraps web fetch calls to request Cloudflare's Markdown for Agents format, returning markdown instead of HTML when the server supports it. Tracks token savings via the x-markdown-tokens response header. Falls back to HTML for sites that don't support the format.
Reduces token consumption by ~80% on supported sites compared to fetching raw HTML, which directly lowers API costs and fits more content into context windows.
Когда его использовать
- Fetching documentation pages with reduced token overhead
- Scraping Cloudflare-hosted sites in agent pipelines
- Replacing axios/fetch calls project-wide with a drop-in optimized version
- Logging token savings across multiple web requests
- Reading web content in LLM context windows more efficiently
Пример сценария
User asks: fetch the content of https://example.com and summarize it
- 1
Call optimizedFetch with the target URL
- 2
Set Accept header to 'text/markdown, text/html'
- 3
Check response Content-Type to determine format
- 4
If text/markdown, use result.markdown directly; otherwise fall back to result.html
- 5
Log token savings from x-markdown-tokens header if present
Markdown content of the page with tokensSaved value reported
Установка
clawhub install howtimeschange/markdown-fetchВыполните команду CLI или вставьте промпт в Claude Code, Cursor, Codex — любой кодинг-агент, которым пользуетесь.