Feature Request: Personal Goggle

Hey,
I like the ideas of goggles, but I _currently_ find the hurdle for creating a custom search experience too high.

## Proposal
A personal Goggle that resides inside the Browser state (or synced with other devices.), that can be updated without ever touching it directly and writing goggle syntax.

## Example
1. You search for something
2. You see results from a news outlet you don't trust.
3. You can _downrank_ it right there next to the search result, adding a `downrank` entry to your personal goggle.

That's it! No need to even know about goggles.

## UX
For improved UX there could be several Buttons like:
`Hide example.org from my results.` → add `$discard,site=example.org` to personal goggle.
`Show more results from example.org` → add `$boost,site=example.org` to personal goggle.
`Show less results from example.org` → add `$downrank,site=example.org` to personal goggle.
(And a lot more, but I think I made my point.)

Even further, it could suggest other discoverable goggles.
> This site appears in the `Copycats removal` Goggle. Do you want to add that Goggle?

## Benefits
1. Easy to discover, understand and use by new or less tech-savvy users.
2. Less friction to maintain your personal goggle, even if you are proficient.
2. Improved privacy, as the goggle is not visible for everyone who knows the URL.

---

Other sites like YouTube already offer options to click on the three dots next to a video and mark it as _Not interested_ or _Don't recommend videos from this channel_. But they don't allow you to see, modify, or share your profile, or enrich it with profiles/goggles from others.

The issue describes a broad personal Goggle feature with browser state, syncing, result controls, and discoverable Goggles, but names no files, tests, or entry points. First map the existing search-result and Goggle configuration flows, then define a narrower implementation and tests before starting; done should include the agreed user actions and persistence behavior.

Written by the indexing model from the issue text.