Enhancement: URL fetcher has no timeout, redirect limit, or max body size

[RobertLD]

The fetchAndConvert() function in src/core/url-fetcher.ts calls fetch() with no:

1. Timeout — a slow server can hang the process forever
2. Redirect limit — could follow infinite redirects
3. Max body size — a 10GB HTML page would be downloaded entirely into memory
4. robots.txt respect — should check before fetching

Should add AbortSignal.timeout(30_000), validate content-length header, and cap the body read.

[RobertLD]

Fixed in commit 926dc80. See the commit message for full details on all fixes.

This issue was resolved as part of a comprehensive fix addressing all 23 open issues.