feat: add built-in free AI service with multi-capability support

[jetyu]

• Introduced a new built-in AI source named 'Snaptium AI' with capabilities for chat, embedding, and reranking.
• Implemented a shared constant for the built-in AI source ID, name, base URL, and model mappings.
• Ensured the built-in source is immutable and cannot be edited or deleted by users.
• Added functionality to fetch a public API token securely and cache it for the Main process lifetime.
• Implemented consent handling for the first use of the built-in AI service, requiring user confirmation before sending data.
• Updated settings normalization to automatically select the built-in source when no user-managed sources exist.
• Enhanced error handling to provide explicit feedback on built-in service failures without switching providers.
• Added comprehensive unit tests to validate the built-in AI service functionality and settings normalization.

Summary

Type of Change

• Bug fix
• Feature
• Documentation
• Refactor
• Build, CI, or release
• Other

Related Issues

Closes #

Changes

Testing

• npm run build:main
• npm run build:preload
• npm run typecheck
• npm run lint
• npm run test:unit
• Not run

Checklist

• I kept this pull request focused and avoided unrelated refactors.
• I followed the Electron layer boundary: Renderer → Preload/Bridge → Main.
• I validated IPC input at the boundary when applicable.
• I avoided any, @ts-ignore, and unnecessary @ts-expect-error.
• I updated Simplified Chinese locale files for new or changed user-visible UI text.
• I updated documentation when behavior, setup, or user-facing functionality changed.
• I considered privacy, security, and user-data impact.