- Add embed_query() default to DenseEmbedderBase (delegates to embed())
- GeminiDenseEmbedder overrides embed_query() using _query_config
  (RETRIEVAL_QUERY); embed() uses _index_config (RETRIEVAL_DOCUMENT)
- Update hierarchical_retriever + memory_deduplicator to call embed_query()
- Deprecate task_type config field (still accepted, no validation error)
- Add enable_multimodal: bool = False flag; supports_multimodal reflects it
- Add embed_multimodal_batch / async_embed_multimodal_batch to base class
- Add Gemini async_embed_multimodal_batch override (anyio semaphore)
- Rewrite embed_multimodal: parts API + pdfminer PDF guard (gated by flag)
- Fix PR volcengine#607 issues #1, #2, #4, #6

…n queuefs/

- Create openviking/storage/queuefs/embedding_handler.py with EmbeddingHandler
  (same logic, corrected class name + docstring)
- Replace TextEmbeddingHandler class body in collection_schemas.py with
  import + backward-compat alias TextEmbeddingHandler = EmbeddingHandler
- Update queue_manager.py to import EmbeddingHandler directly from queuefs
- Fixes PR volcengine#607 issue #3

… embedding

- Add ContentPart = Union[str, ModalContent] type alias
- Add parts: Optional[List[ContentPart]] field to Vectorize
- Add get_parts(): returns parts if set, else builds [text, media] from legacy fields
- Add multi-part integration tests (TestGeminiE2EMultipartEmbedding)
- Fixes PR volcengine#607 issue #1 (multi-part sequences)

…gine#607 review)

Merge 17 upstream commits (feat/fix: trace metrics, multi-agent isolation,
vikingdb TUI, rate-limit simplification, session fixes, etc.) and resolve
conflicts in embedding_msg, embedding_msg_converter, collection_schemas,
pyproject.toml, and uv.lock.

Scope reduction (defers multimodal pipeline to April when add-resource
pipeline is ready):
- Remove multimodal dispatch from embedding_utils.vectorize_file; media
  files always fall back to text/summary (resolves r2936557627: removes
  hardcoded "gemini" string check; resolves r2936556516: no PDF 6-page
  limit concern)
- Drop media_uri/media_mime_type from EmbeddingMsg; add telemetry_id
  from upstream; preserve id through to_dict/from_dict round-trip
- Drop media fields from EmbeddingMsgConverter; pass telemetry_id
- Remove multimodal on_dequeue path from TextEmbeddingHandler; adopt
  upstream's telemetry-tracked text-only path
- Add TODO in GeminiDenseEmbedder.embed_multimodal for parts-list
  aggregation (resolves r2936536550 — deferred to April)

GeminiDenseEmbedder class retains supports_multimodal + embed_multimodal
for future activation once Vectorize.parts is supported end-to-end.

… image/video/audio/PDF) (#607)"

This reverts commit 95bd197.

… image/video/audio/PDF) (#607)" (#703)

This reverts commit 95bd197.

…ideo/audio/PDF) (volcengine#607)

* feat(embedder): Gemini Embedding 2 multimodal support (text + image/video/audio/PDF)

Native text + multimodal (image, video, audio, PDF) embedding via `gemini-embedding-2-preview` (google-genai 1.67.0). Additive provider pattern — Volcengine remains the default; Gemini is opt-in via `provider: "gemini"` in `ov.conf`.
    - **Model**: `gemini-embedding-2-preview`
    - **Input**: text, image, video, audio, PDF (17 MIME types)
    - **Output dimension**: 128–3072 (default: **3072**, recommended: 768 / 1536 / 3072)
    - **Input token limit**: **8,192 tokens**
    - **Supported MIME types**: `image/jpeg`, `image/png`, `image/gif`, `image/webp`, `audio/mpeg`, `audio/mp3`, `audio/wav`, `audio/ogg`, `audio/flac`, `video/mp4`, `video/mpeg`, `video/mov`, `video/avi`, `video/webm`, `video/wmv`, `video/3gpp`, `application/pdf`
- Gemini Embedding 2 Multimodal Support: Introduced a new GeminiDenseEmbedder to support native text and multimodal (image, video, audio, PDF) embedding using the gemini-embedding-2-preview model. This is an opt-in provider via configuration.
- Extended Queue Pipeline for Multimodal Content: The EmbeddingMsg now carries media_uri and media_mime_type to facilitate multimodal content processing. The TextEmbeddingHandler.on_dequeue() method was updated to read raw bytes from viking_fs and call embed_multimodal() when applicable.
- End-to-End Configuration and Security: The EmbeddingConfig now registers the 'gemini' provider with a task_type field. A critical security validation was added to ensure media_uri matches context_data['uri'] before file reads, preventing forged queue messages from accessing arbitrary files. If validation fails or multimodal embedding fails, it falls back to text embedding.
- Multimodal Content Representation: A new ModalContent dataclass was introduced to represent media references, including MIME type, URI, and optional raw data, enabling the Vectorize object to encapsulate both text and media for embedding

* feat: Add asynchronous batch embedding with concurrency control to Gemini embedder.

* Reduce scope to use GeminiDenseEmbedder as only text embed

… image/video/audio/PDF) (volcengine#607)" (volcengine#703)

This reverts commit 95bd197.