What problem does it solve? Reading and writing CSV, TSV, and other delimited files in .NET often involves tradeoffs between allocation overhead, parsing correctness, and configuration control. This Skill provides guidance for using the Sep library to handle separated-value data with explicit reader/writer options, low-allocation span-based APIs, and support for header inference, trimming, unescaping, and async pipelines. ## Core Features & Use Cases - High-performance reading: Parse delimited files with Sep.Reader, access columns as ReadOnlySpan<char>, and convert values with Parse<T> only when needed. - Controlled writing: Generate output with SepWriterOptions, round-trip input specs via reader.Spec.Writer(), and control header, escaping, and column-count behavior. - Async and parallel workflows: Use async readers/writers on .NET 9+ and apply ParallelEnumerate for CPU-heavy transformations after benchmarking. - Use Case: An ETL pipeline reads a large CSV file, transforms numeric columns, and writes normalized output while preserving the original separator and culture settings. ## Quick Start Ask the AI to add the Sep NuGet package to your .NET project and write code that reads a CSV file with Sep and writes the transformed rows to a new file.