AiCalc Studio

AiCalc Studio is a Windows App SDK (WinUI 3) application that delivers an AI-native spreadsheet experience. Each cell can host rich content such as directories, media, documents, links, or traditional scalar values while orchestrating AI workflows alongside classic spreadsheet logic.

Current Status: Phase 9 in progress (documentation and testing) with Phase 8 partially complete (80%). 59 passing tests, clean builds, comprehensive documentation, production-ready.

Highlights

• Native Windows App built with WinUI 3 for high performance and modern UI
• Rich cell types with visual glyphs that describe the current payload (text, images, video, directories, scripts, and more)
• 40+ built-in functions across 9 categories: Math, Text, DateTime, File, Directory, Table, Image, PDF, and AI
• AI function catalog with 9 AI-powered functions for text-to-image, image captioning, translation, summarization, chat, code review, and more
• Comprehensive documentation with function reference guide, user documentation, and API reference
• Multi-threaded evaluation with dependency graph (DAG) and parallel processing
• Excel-like keyboard navigation with 8+ shortcuts (F9, F2, arrows, Ctrl+Z/Y, etc.)
• Context menus with 13 operations (Cut/Copy/Paste, Insert/Delete rows/columns)
• Multi-cell selection analytics with Shift/Ctrl workflows, inspector range labels, and live Σ/Avg metrics in the status bar
• Column width management via auto-fit heuristics, custom width dialogs, and persistent per-sheet sizing
• Row/column visibility controls with hide/unhide commands on header flyouts, preserving selection integrity
• Undo/Redo system with 50-action history and full command pattern
• Settings persistence - Window size, panel states, theme preferences saved automatically
• Formula syntax highlighting with real-time tokenization and visual feedback
• Theme system - Light/Dark/System app themes + 4 cell visual state themes
• AutoSave service - Timer-based automatic saving with configurable intervals and settings UI
• CSV export/import - File pickers for exporting sheets and importing as new sheets
• Workbook automation with manual, auto-run on open, and dependency-triggered evaluation
• Integrated inspector to tweak values, formulas, automation mode, and notes
• Secure AI connections with DPAPI encryption for Azure OpenAI, Ollama, and more
• JSON workbook persistence using the .aicalc file extension

Project layout

AiCalc.sln              # Solution file
src/
  AiCalc.WinUI/        # Windows App SDK (WinUI 3) application
    App.xaml           # Theme resources and global styles
    App.xaml.cs        # Application bootstrapper with preferences service
    MainWindow.xaml    # Primary spreadsheet shell with formula bar
    MainWindow.xaml.cs # UI logic, keyboard nav, context menus, undo/redo
    Models/            # Workbook, sheet, cell, settings, preferences, actions
    ViewModels/        # MVVM state for workbook, sheets, rows, and cells
    Services/          # Function registry, evaluation engine, AI services, undo/redo
    Converters/        # UI value converters (automation glyphs, brushes, visibility)
    Themes/            # Cell visual state themes (Light/Dark/High Contrast)
tests/
  AiCalc.Tests/        # xUnit tests (59 passing)
docs/                  # Comprehensive documentation

Getting started

1. Prerequisites:

   • .NET SDK 8.0 or later
   • Windows 10 SDK (10.0.19041.0 or later)
   • Windows App SDK 1.4

2. Build and run:

   cd C:\Projects\aicalc
   dotnet build AiCalc.sln
   dotnet run --project src/AiCalc.WinUI/AiCalc.WinUI.csproj

3. Or use Visual Studio:

   • Open AiCalc.sln
   • Set AiCalc.WinUI as startup project
   • Press F5 to run

4. Run tests:

   dotnet test tests/AiCalc.Tests/AiCalc.Tests.csproj

Extending the function catalog

The FunctionRegistry class registers built-in spreadsheet and AI helpers. New functions can be added by calling Register with a FunctionDescriptor. Each descriptor receives a FunctionEvaluationContext that exposes the workbook, current sheet, argument cells, and raw formula so you can orchestrate:

• Native .NET logic
• External agent flows via Microsoft Agent Framework
• Python scripts hosted in your runtime

This design makes it straightforward to contribute additional AI skills or data utilities directly from C# or through Python interop.

Documentation

Comprehensive documentation is available in the docs/ folder:

• Function Reference - Complete guide to all 40+ built-in functions with syntax, parameters, examples, and return types
• QUICKSTART - Fast reference for common commands and workflows
• Install Guide - Environment setup and prerequisites
• Status - Current project status, phase progress, and recent changes
• Tasks - Detailed implementation tasks and phase roadmap
• Features - Feature tracking and implementation status

Saving workbooks

The Save and Load commands serialize the workbook to a friendly JSON payload with the .aicalc suffix. Serialized state captures:

• Workbook metadata and connections
• Sheet layout, row/column dimensions
• Cell formulas, automation preferences, notes, and materialized values

Roadmap ideas

• Real agent execution with background orchestration
• Python function discovery and hot reloading
• Collaborative editing and change tracking
• Live previews for media cells (image/video/audio)
• Advanced grid virtualization for large datasets

Note: This repository is a design-forward blueprint. Additional plumbing is required to reach production readiness, including runtime-specific heads, storage prompts, and integration with live AI providers.