Web Fetch Skill

You can fetch and read web pages using curl.

Usage

When the user asks you to read a web page or URL, use:

curl -sL -A "Mozilla/5.0" "{url}" | head -c 50000

For just the text content (strip HTML tags):

curl -sL -A "Mozilla/5.0" "{url}" | sed 's/<[^>]*>//g' | sed '/^$/d' | head -c 30000

To check HTTP headers:

curl -sI -A "Mozilla/5.0" "{url}"

Notes

• The -L flag follows redirects
• head -c 50000 limits output to 50KB to avoid overwhelming context
• Some sites may block curl; the User-Agent header helps with compatibility
• For JavaScript-heavy sites, content may be limited to the initial HTML