Feature: Semantic Search with Embeddings

[avikeid2007]

Summary

Add embedding-based semantic search across chat history and documents using LLamaSharp's embedding support.

Motivation

Current search is keyword-based. Semantic search finds related content by meaning, dramatically improving the search experience across chat history and documents.

Detailed Requirements

Service: EmbeddingService.cs

public interface IEmbeddingService
{
    Task<float[]> GenerateEmbeddingAsync(string text);
    Task IndexMessageAsync(string sessionId, string messageContent, string messageId);
    Task<List<SearchResult>> SemanticSearchAsync(string query, int topK = 10);
    Task RebuildIndexAsync(IProgress<int> progress = null);
}

public class SearchResult
{
    public string MessageId { get; set; }
    public string SessionId { get; set; }
    public string Content { get; set; }
    public float Score { get; set; }  // Cosine similarity
}

• Use LLamaEmbedder from LLamaSharp to generate embeddings
• Store embeddings in SQLite as BLOB
• Cosine similarity search for finding related content
• Index: chat messages, documents, notes (when those features exist)

Database (DatabaseService.cs)

Add Embeddings table:

CREATE TABLE IF NOT EXISTS Embeddings (
    Id TEXT PRIMARY KEY,
    SourceType TEXT NOT NULL,   -- 'message', 'document', 'note'
    SourceId TEXT NOT NULL,
    Content TEXT NOT NULL,
    Embedding BLOB NOT NULL,
    CreatedAt TEXT NOT NULL
);

UI

• Enhanced search in chat history sidebar with "Semantic" toggle button
• Search results ranked by relevance score (show score as percentage)
• Highlight matching content in search results

Background Indexing

• Index new messages automatically after each conversation
• Use a background thread/task to avoid impacting chat performance
• Show indexing progress in status bar

Files to Create/Modify

• [NEW] Services/EmbeddingService.cs + IEmbeddingService.cs
• [NEW] Models/SearchResult.cs
• [MODIFY] Services/DatabaseService.cs - add Embeddings table
• [MODIFY] ViewModels/ChatViewModel.cs - integrate semantic search
• [MODIFY] Views/ChatView.xaml - semantic search toggle in history panel

Acceptance Criteria

• Embeddings generated for chat messages using LLamaSharp
• Semantic search returns relevant results by meaning, not just keywords
• Search results ranked by cosine similarity score
• Background indexing does not impact chat performance
• Works with any loaded GGUF model that supports embeddings