[geaflow/ai-memory] Implement `ChunkVectorIndex`

Priority: P0
Difficulty: Intermediate

Context: HugeGraph-AI-style RAG retrieves document chunks. geaflow-ai currently indexes `GraphEntity`, not document chunks.

Scope:

- Add chunk vector index built on `VectorStore`.
- Store `chunk_id`, source span, embedding, model metadata, and text hash.
- Add tests for insert, query, delete marker, restart, and source span return.

Constraints:

- Do not call remote embedding services in tests.
- Use deterministic fake embeddings.
- Do not return raw full document when only source span is needed.

Acceptance Criteria:

- Query vector returns chunk hits with score and `source_ref`.
- Deleted chunks do not appear after searchable watermark.

Suggested paths:

- `geaflow-ai/src/main/java/org/apache/geaflow/ai/index/chunk`

Start in geaflow-ai/src/main/java/org/apache/geaflow/ai/index/chunk and trace the existing GraphEntity indexing flow and VectorStore interface. Add tests for insert, query, delete markers, restart behavior, deterministic fake embeddings, and source-span returns. Done means queries return chunk hits with scores and source_ref, while deleted chunks disappear after the searchable watermark.

Written by the indexing model from the issue text.