UI Features Guide

Complete guide to all features available in the Scrapalot web application.

Chat Interface

Real-time AI Conversations

Chat Flow Decision Tree:

Features:

• Streaming Responses - See answers as they're generated in real-time
• Citation Tracking - Every answer shows which documents were used
• Thinking Sections - Watch the AI reason through complex questions (optional)
• Message History - Scroll through past conversations
• Edit Messages - Modify previous messages and regenerate responses
• Multi-turn Conversations - Ask follow-up questions with full context

Chat Toolbar

Located above the chat input, the toolbar provides quick access to:

Model Selector:

• Choose from OpenAI, Anthropic, Google, Ollama, and custom providers
• Filter by model type (chat, vision, audio)
• See context window sizes and capabilities

Collection Selector:

• Select which document collections to search
• Multi-select for searching across collections
• Create new collections on the fly

Prompt Templates:

• Quick access to saved prompts
• Create custom templates for repetitive tasks
• Organize by category

File Attachments:

• Upload documents directly into conversation
• Support for PDF, images, text files
• Attach to specific messages

Advanced Settings:

• Web search toggle - Include real-time web results
• Similarity threshold - Control relevance (0-1)
• Number of chunks - How many document sections to use (1-50+)
• Deep research mode - Multi-step research with visualization

Document Management

Knowledge Stack Dialog

The Knowledge Stack is your document library.

Document Processing Pipeline:

Features:

• Multi-file Upload - Drag & drop multiple files at once
• Processing Tracker - Real-time status of embedding generation
• Collection Management - Create, rename, delete collections
• Document Search - Find documents by name
• Storage Quota Display - Track your workspace storage usage
• Resizable Sidebar - Adjust panel width for better visibility

Supported File Types:

• PDF documents
• EPUB books
• Microsoft Word (.docx, .doc)
• Text files (.txt, .md)
• CSV files
• And more...

Document Viewers

PDF Viewer:

• Opens in right-side drawer
• Read while chatting
• Jump to citations automatically
• Zoom, rotate, download controls

EPUB Viewer:

• Dedicated book reader
• Table of contents navigation
• Bookmarks and highlights
• Reading progress tracking

Settings & Configuration

8 Settings Tabs

Access via the settings icon in the sidebar:

1. General Settings

Theme & Appearance:

• Color Mode - Light, dark, or system-based
• Accent Color - 6 color options (gray, blue, green, red, violet, orange)
• Font Settings - Family, size, code theme
• Language - English, Croatian, and more

Display Options:

• Toggle link rendering
• Show/hide AI reasoning indicators
• Auto-save indicator
• Custom rendering modules

2. Account Tab

Profile Management:

• Edit name, username, email
• Upload profile picture
• Delete profile picture

Security:

• Change password
• Account deletion option

Subscription:

• Current plan details
• Storage quota visualization
• Upgrade options
• Usage statistics

3. Remote Providers Tab

Configure cloud AI providers:

Provider	Description	Setup
OpenAI	Industry-leading AI models	API key required
Anthropic	Advanced reasoning models	API key required
Google	Multimodal AI models	API key required
Azure	Azure OpenAI service	Endpoint + key
DeepSeek	Cost-effective models	API key required
Mistral AI	European AI provider	API key required
xAI	Cutting-edge AI models	API key required
Cohere	Enterprise AI models	API key required
OpenRouter	Multi-provider access	API key required
HuggingFace	Open model inference	API token
Custom	OpenAI-compatible endpoints	Base URL + key

Features:

• Add multiple providers
• Test API connections
• Select which models to enable
• Manage API keys securely

4. Local AI Tab

Run models on your own hardware:

Ollama Integration:

• Status monitoring (running/stopped)
• Model management
• Download models from Ollama library
• Configure host and port

Hardware Detection:

• GPU information display
• CUDA/Metal capability detection
• Memory availability
• Recommended GPU layers

Featured Models:

• Llama 3.1, Llama 3.2
• Mistral, Mixtral
• DeepSeek models
• And more...

Deploy Models:

• One-click model deployment
• Configure context length
• Set GPU layers
• Temperature and sampling controls

5. Prompts Tab

Template Management:

• Create custom prompt templates
• Organize by category
• Activate/deactivate templates
• Quick access in chat toolbar

Use Cases:

• Code review templates
• Document summarization
• Translation prompts
• Custom instructions

6. Documents Tab

Configure document processing:

• Choose embedding model
• Set default collection
• Adjust chunking strategy
• Configure vectorization settings

7. Workspaces Tab

Workspace Management:

• View all workspaces
• Switch between workspaces
• Create new workspaces
• Configure workspace settings

Team Features:

• View team members
• Invite teammates
• Manage user roles (Owner, Editor, Viewer)
• Remove team members

8. Service Tab

Backend Configuration:

• Monitor service status
• View connection health
• Advanced service settings
• Diagnostics tools

Collaboration Features

Team Management

Collaboration Use Cases:

User Roles:

Role	Can View	Can Edit	Can Upload	Can Invite
Owner	✅	✅	✅	✅
Editor	✅	✅	✅	❌
Viewer	✅	❌	❌	❌

Notes & Comments

Collaborative Notes Editor:

• Rich text formatting (bold, italic, headings, lists)
• Insert links and images
• Code blocks with syntax highlighting
• Real-time collaborative editing
• Comments panel for team discussions
• Version history
• Share notes with team

Slash Commands:

• /heading1, /heading2, /heading3
• /bullet, /number
• /code, /quote
• /image, /link

Advanced Features

Deep Research

Multi-step Research Decision Flow:

Configuration:

• Breadth - Number of exploration paths (1-10)
• Depth - Research iterations (1-5)
• Visualization - See research steps in real-time
• Source Tracking - View all sources used

Use Cases:

• Complex research questions
• Literature reviews
• Market analysis
• Comprehensive reports

Web Search Integration

Features:

• Toggle web search on/off per query
• Real-time web results
• Source attribution
• Combine documents + web knowledge

When to Use:

• Current events and news
• Latest information not in documents
• Fact-checking and verification
• Supplementing document knowledge

Mobile & Responsive Design

Mobile Features:

• Full-screen interface on mobile
• Touch-optimized controls
• Drawer-based navigation
• Responsive sidebars collapse automatically

Tablet Features:

• Portrait mode optimizations
• Sidebar as full-screen menu
• Touch-friendly buttons

Keyboard Shortcuts

Speed up your workflow with keyboard shortcuts:

Action	Windows/Linux	Mac
New conversation	Ctrl+N	⌘+N
Open settings	Ctrl+,	⌘+,
Focus chat input	/	/
Submit message	Enter	Enter
New line in message	Shift+Enter	Shift+Enter
Search collections	Ctrl+K	⌘+K

Performance Features

Optimizations Built-in:

• Request Deduplication - Prevent duplicate API calls (5-second window)
• Response Caching - Cache GET responses (1-minute TTL)
• Message LRU Cache - Keep last 10 sessions in memory
• Lazy Loading - Load workspace data only when needed
• Code Splitting - Fast initial page load

Accessibility

Features:

• Keyboard navigation support
• Screen reader compatible
• High contrast themes
• Resizable text
• Semantic HTML
• ARIA labels

Next Steps

Learn More

• Uploading Documents - Detailed document upload guide
• Asking Questions - Get better answers
• Collections - Organize your knowledge
• FAQ - Common questions