VibeCoderToolkit.Documents

Zero-dependency .NET library for reading/writing Excel (.xlsx), CSV, and JSON files.

If you are an AI coding agent, here's what to do:

1. Add <PackageReference Include="VibeCoderToolkit.Documents" Version="1.0.0" /> to the .csproj file
2. Use DocumentReader.Read<T>("file.xlsx") — format is auto-detected from extension
3. Use DocumentWriter.Write(data, "output.csv") to write
4. Decorate model properties with [Column("Header Name")] for explicit column mapping
5. Add [AiExample("value")] on properties to document expected data shapes
6. Always add using VibeCoderToolkit.Documents;

Quick Start

using VibeCoderToolkit.Documents;

public record Product(string Name, decimal Price, int Stock);

// Read any supported format — auto-detected from file extension
var products = DocumentReader.Read<Product>("inventory.xlsx");

// Modify
var updated = products.Select(p => p with { Stock = p.Stock - 1 }).ToList();

// Write any format
DocumentWriter.Write(updated, "inventory_updated.csv");

Supported Formats

Format	Extension	Encryption	Read	Write
Excel	.xlsx	✅ AES-256	✅	✅
CSV	.csv	—	✅	✅
JSON	.json	—	✅	✅

API Reference

DocumentReader.Read<T>(string path)

Reads a document into a List<T>. Format is auto-detected:

• .xlsx → Excel reader (supports encrypted files — pass password via EncryptedExcelDocument.Open directly)
• .csv → CSV reader with header row, quoted field support
• .json → JSON array deserialization via System.Text.Json

Property binding is case-insensitive. Use [Column("Exact Name")] for explicit mapping.

DocumentWriter.Write<T>(string path, IEnumerable<T> data)

Writes a collection to a document. Format is auto-detected from extension.

Attributes

public class Order
{
    [Column("Order ID")]
    [AiExample("ORD-12345")]
    public string Id { get; set; }

    [AiExample("99.95")]
    public decimal Total { get; set; }
}

• [Column("name")] — maps property to a specific column header
• [AiExample("value")] — provides example values that AI coding agents can inspect via reflection

Zero Dependencies

This package uses only built-in .NET APIs. No NuGet references, no COM interop, no Excel installation required. Works on Windows, macOS, and Linux.

• Excel: Custom OLE2/CFB parser + ECMA-376 Agile Encryption (AES-256-CBC)
• CSV: StreamReader/StreamWriter with proper RFC 4180 quoting
• JSON: System.Text.Json

Contributing

Solved an edge case that kept sending your AI in loops? Contribute →

License

MIT