feat(amazon-bedrock): add guardrail content support for text and image parts (#14302)

## Background

The Amazon Bedrock Converse API supports [guardrail content
blocks](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_GuardrailConverseContentBlock.html)
that allow marking text and image parts for guardrail evaluation with
optional qualifiers (`grounding_source`, `query`, `guard_content`). The
existing `BedrockGuardrailConverseContentBlock` type was untyped
(`guardContent: unknown`) and there was no way to opt into this behavior
from the AI SDK.

[Documentation on the Bedrock
Feature](https://docs.aws.amazon.com/bedrock/latest/userguide/guardrails-use-converse-api.html)

## Summary

- Added `BedrockGuardrailTextBlock` type extending `BedrockTextBlock`
with optional `qualifiers` array.
- Typed `BedrockGuardrailConverseContentBlock.guardContent` as a union
of text and image variants instead of `unknown`.
- Added Zod schemas (`bedrockTextPartProviderOptions`,
`bedrockImagePartProviderOptions`) for validating provider options, used
via `parseProviderOptions` consistent with existing patterns.
- Users can now set `providerOptions.bedrock.guardContent: true` on text
and image parts to wrap them as `guardContent` blocks, with optional
`guardContentQualifiers` for text parts.
- Added 8 tests covering text guard content (with/without qualifiers,
false, no options) and image guard content (true, false, no options).

## Checklist

- [x] Tests have been added / updated (for bug fixes / features)
- [x] Documentation has been added / updated (for bug fixes / features)
- [x] A _patch_ changeset for relevant packages has been added (for bug
fixes / features - run `pnpm changeset` in the project root)
- [x] I have reviewed this pull request (self-review)

Co-authored-by: Aayush Kapoor <aayushkapoor34@gmail.com>