Auto-embedding of text fields into vector fields

New Feature / Enhancement Checklist

• Report security issues confidentially.
• Any contribution is under this license.
• Before posting search existing issues.

Current Limitation

Even with a vector field (#10605), developers have to generate embeddings themselves in Cloud Code and keep them in sync whenever the source text changes. That is boilerplate everyone building semantic search would have to repeat.

Feature / Enhancement Description

Declare in the schema that a string field gets embedded into a vector field on save, e.g. description -> descriptionEmbedding. Parse Server keeps the two in sync automatically via an internal save hook, using a server configured embedding provider (#10608).

Example Use Case

1. Configure the mapping on a class in the schema
2. Save objects normally, no embedding code anywhere
3. Semantic search over that class just works via $nearVector

Alternatives / Workarounds

DIY afterSave trigger that calls an embedding API and writes the vector back. Works, but every app reimplements the same sync, dirty checking and failure handling.

3rd Party References

• Supabase automatic embeddings
• Firebase "vector embeddings with Firestore" extension

Depends on #10605

Start by reading the related vector-field issue (#10605) and embedding-provider issue (#10608), then trace Parse Server's schema handling and internal save-hook entry points. Define how a schema mapping such as description -> descriptionEmbedding is configured, synchronized on save, and handles failures; done means semantic search via $nearVector works without application Cloud Code.