[Feature]: Integrate SeekDB as a Database Component

**AgentScope is an open-source project. To involve a broader community, we recommend asking your questions in English.**

## Feature Request: Integrate SeekDB as a Database Component

**Is your feature request related to a problem? Please describe.**

AgentScope currently lacks a unified database solution that can efficiently handle the diverse data requirements of AI agent applications.

**Describe the solution you'd like**

I propose integrating [SeekDB](https://github.com/oceanbase/seekdb) as an optional database component for AgentScope. SeekDB is an AI-native, open-source database that unifies vector, text, structured, and semi-structured data within a single engine.

### Key Features:

1. **Hybrid Search**
- Combines vector-based semantic retrieval, full-text keyword search, and scalar filtering in a single query
- Supports query reranking with multiple strategies (weighted scores, reciprocal rank fusion, LLM-based reranking)

2. **Unified Data Model**
- Handles relational data with standard SQL
- Vector search with multiple distance metrics (cosine, Euclidean, inner product)
- Full-text search with BM25 ranking
- JSON and GIS data support

3. **In-Database AI Functions**
- `AI_EMBED`: Convert text to embeddings directly in SQL
- `AI_COMPLETE`: Generate text using chat/completion models
- `AI_RERANK`: Rerank search results

4. **Lightweight Deployment**
- Embedded mode for local development
- Client/server mode for production
- Minimal resource requirements (1 CPU core, 1GB memory)
- MySQL protocol compatible

### Integration Approach:

The integration would provide a `SeekDBStorage` class that can be used as a storage backend for AgentScope:

```python
from agentscope.storage import SeekDBStorage

# Simple usage
storage = SeekDBStorage(
connection_string="mysql://127.0.0.1:2881/seekdb",
enable_ai_functions=True
)

# Use for vector storage
storage.add_vectors(embeddings, metadata)

# Hybrid search
results = storage.hybrid_search(
vector_query=query_embedding,
text_query="keyword search",
filters={"category": "document"}
)
```

**Describe alternatives you've considered**

1. **Multiple Database Services**: Using separate vector DB, text search, and relational DB services would work but increases architectural complexity and operational overhead.

2. **Existing Vector Databases**: While specialized vector databases (e.g., Pinecone, Weaviate) exist, they don't provide the unified hybrid search and in-database AI capabilities that SeekDB offers.

3. **Building Custom Integration**: This would require significant development effort and wouldn't leverage SeekDB's proven, production-ready features.

**Additional context**

### Why SeekDB is a Good Fit:

1. **AI-Native Design**: Built specifically for AI applications with in-database AI functions
2. **Open Source**: Apache 2.0 license, actively maintained by OceanBase
3. **Unified Solution**: Single database for all data types, reducing architecture complexity
4. **MySQL Compatible**: Easy integration with existing tools and ecosystems
5. **Lightweight**: Can run embedded or as a server, suitable for various deployment scenarios

### Resources:

- **SeekDB GitHub**: https://github.com/oceanbase/seekdb
- **Documentation**: Available on GitHub repository
- **License**: Apache 2.0