WIP: LLM Integration

db/migrations/0229__note_embeddings.sql

@@ -0,0 +1,24 @@
+CREATE TABLE IF NOT EXISTS "note_embeddings" (
+    embeddingId TEXT NOT NULL PRIMARY KEY,
+    noteId TEXT NOT NULL,
+    model TEXT NOT NULL,
+    vector BLOB NOT NULL,
+    dimension INTEGER NOT NULL,
+    utcDateCreated TEXT NOT NULL,
+    utcDateModified TEXT NOT NULL,
+    isDeleted INTEGER NOT NULL DEFAULT 0,
+    FOREIGN KEY(noteId) REFERENCES notes(noteId)
+);
+
+CREATE INDEX IF NOT EXISTS IDX_note_embeddings_noteId ON note_embeddings (noteId);
+CREATE INDEX IF NOT EXISTS IDX_note_embeddings_model ON note_embeddings (model);
+
+INSERT INTO options (name, value, isSynced, utcDateModified)
+VALUES 
+('llmEnabled', 'false', 1, datetime('now')),
+('llmProvider', 'ollama', 1, datetime('now')),
+('llmModel', 'llama2', 1, datetime('now')),
+('llmBaseUrl', 'http://localhost:11434', 1, datetime('now')),
+('llmMaxTokens', '2000', 1, datetime('now')),
+('llmTemperature', '0.7', 1, datetime('now')),
+('llmApiKey', '', 1, datetime('now')); 

docs/llm_integration.md

@@ -0,0 +1,108 @@
+# LLM Integration in Notes
+
+Notes now includes built-in support for Large Language Models (LLMs), allowing you to chat with your notes and generate embeddings for semantic search. The integration is designed to be provider-agnostic, supporting various LLM providers such as OpenAI, Ollama, and others.
+
+## Features
+
+1. Chat with your notes
+   - Access the chat interface from the right panel
+   - Chat uses the context of your current note
+   - Supports markdown formatting in responses
+
+2. LLM Provider Configuration
+   - Configure your preferred LLM provider in Settings -> Appearance -> LLM Settings
+   - Supported providers:
+     - Ollama (default, requires local installation)
+     - OpenAI (requires API key)
+
+3. Note Embeddings
+   - Automatically generates embeddings for your notes
+   - Embeddings are used for semantic search and finding similar notes
+   - Embeddings are stored locally in the SQLite database
+
+## Setup
+
+### Using Ollama (Default)
+
+1. Install Ollama from https://ollama.ai/
+2. Start the Ollama service
+3. In Notes, go to Settings -> Appearance -> LLM Settings
+4. Select "Ollama" as the provider
+5. Set the model (default: "llama2")
+6. Configure the base URL if needed (default: "http://localhost:11434")
+
+### Using OpenAI
+
+1. Get an API key from OpenAI
+2. In Notes, go to Settings -> Appearance -> LLM Settings
+3. Select "OpenAI" as the provider
+4. Enter your API key
+5. Set the model (default: "gpt-3.5-turbo")
+
+## Usage
+
+### Chatting with Notes
+
+1. Open a note you want to chat about
+2. Click the "Chat" tab in the right panel
+3. Type your message and press Enter or click Send
+4. The LLM will respond using the context from your current note
+
+### LLM Settings
+
+Access LLM settings through:
+1. Settings menu
+2. Appearance section
+3. Click "Configure LLM" button
+
+Available settings:
+- Provider selection
+- Model selection
+- API key (for OpenAI)
+- Base URL (for Ollama)
+- Temperature (controls response randomness)
+- Max tokens (controls response length)
+
+## Technical Details
+
+### Embeddings Storage
+
+Embeddings are stored in the `note_embeddings` table with the following schema:
+```sql
+CREATE TABLE note_embeddings (
+    embeddingId TEXT NOT NULL PRIMARY KEY,
+    noteId TEXT NOT NULL,
+    model TEXT NOT NULL,
+    vector BLOB NOT NULL,
+    dimension INTEGER NOT NULL,
+    utcDateCreated TEXT NOT NULL,
+    utcDateModified TEXT NOT NULL,
+    isDeleted INTEGER NOT NULL DEFAULT 0,
+    FOREIGN KEY(noteId) REFERENCES notes(noteId)
+);
+```
+
+### Synchronization
+
+Embeddings are considered local to each client and are not synchronized between devices. This is because:
+1. Embeddings can be regenerated on each client
+2. Different clients might use different LLM providers or models
+3. Reduces sync data size and complexity
+
+### API Endpoints
+
+The following API endpoints are available:
+
+- `POST /api/llm/chat` - Send a chat message
+- `POST /api/llm/embeddings/:noteId` - Generate embeddings for a note
+- `GET /api/llm/config` - Get LLM configuration
+- `PUT /api/llm/config` - Update LLM configuration
+
+## Contributing
+
+The LLM integration is designed to be extensible. To add support for a new LLM provider:
+
+1. Create a new provider class in `src/services/llm/providers/`
+2. Implement the `LLMProvider` interface
+3. Register the provider in `src/services/llm/llm_service.ts`
+