browser: add file download support

[jmandel]

When navigating to a URL that triggers a download (PDF, binary files, etc.), the browser now captures the download and saves it to /tmp/shelley-downloads/ instead of failing.

Changes:

• Configure Chrome's download behavior on browser initialization
• Listen for download events (EventDownloadWillBegin, EventDownloadProgress)
• Use page.Navigate to detect downloads via the isDownload return flag
• Save files with browser-style naming: file.pdf, file (1).pdf, file (2).pdf, etc.
• Return clear message: "File downloaded to: /path/file.pdf (1234 bytes)"

This allows the LLM to:

1. Navigate to pages with content protection (CAPTCHAs, login walls)
2. Interact with the page as needed
3. Download files that are behind protection
4. Use other tools (pdftotext, etc.) to process the downloaded files

The file is saved directly to disk, avoiding the need to base64-encode large files through the LLM's context window.

Fixes #7

[cla-bot]

Thank you for your pull request and welcome to our community. We could not parse the GitHub identity of the following contributors: exe.dev user.
This is most likely caused by a git client misconfiguration; please make sure to:

1. check if your git client is configured with an email to sign commits git config --list | grep email
2. If not, set it up using git config --global user.email email@example.com
3. Make sure that the git commit email is configured in your GitHub account settings, see https://github.com/settings/emails

[cla-bot]

We require contributors to sign our Contributor License Agreement, and we don't have you on file. In order for us to review and merge your code, please contact @crawshaw to get yourself added.

[cla-bot]

We require contributors to sign our Contributor License Agreement, and we don't have you on file. In order for us to review and merge your code, please contact @crawshaw to get yourself added.