feat(appsync-modelgen-plugin): add metrics and usage fields to conversation message model

[pjkroker]

Description of changes

Adds metrics and usage fields to the generated conversation message model in processConversationRoute.

This enables clients to receive token usage metadata (input/output/total tokens) and metrics from Bedrock AI conversation responses, supporting the ability to track and limit token usage per user.

The fields are typed as:

• metrics: AmplifyAIMetrics (nonModel)
• usage: AmplifyAIUsage (nonModel)

Codegen Paramaters Changed or Added

Issue #, if available

aws-amplify/amplify-backend#2355

Description of how you validated changes

• Ran yarn build and yarn test in packages/appsync-modelgen-plugin — all tests pass
• Updated snapshot for appsync-model-introspection-visitor to include the new fields
• graphql-generator test failures are pre-existing on main and unrelated to this change

Checklist

• [ x] PR description included
• [ x] yarn test passes
• E2E test run linked
• [ x] Tests are changed or added
• Relevant documentation is changed or added (and PR referenced)
• Breaking changes to existing customers are released behind a feature flag or major version update
• Changes are tested using sample applications for all relevant platforms (iOS/android/flutter/Javascript) that use the feature added/modified
• Changes are tested on windows. Some Node functions (such as path) behave differently on windows.
• Changes adhere to the GraphQL Spec and supports the GraphQL types type, input, enum, interface, union and scalar types.

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.