text_skill

Overview

This skill transforms, formats, and processes text using reliable patterns for writing, data cleaning, localization, citations, and marketing copy. It collects practical commands and rules for encoding, whitespace normalization, format detection, common traps, and quick text transformations. Use it to prep text for parsing, translation, publishing, or analytics workflows.

How this skill works

The skill inspects file encoding, line endings, delimiters, and character anomalies, then applies deterministic transformations like trimming, collapsing whitespace, and normalizing punctuation. It provides shell-friendly commands and regex snippets to extract emails, URLs, count words, and detect CSV delimiters. The workflow emphasizes verifying encoding first, normalizing text, and running targeted transformations while preserving Unicode correctness.

When to use it

• Preparing text for parsers or ingestion pipelines
• Cleaning datasets before analysis or machine learning
• Normalizing content for localization or translation
• Formatting academic citations and marketing copy consistently
• Detecting delimiters in CSV/TSV imports

Best practices

• Always verify encoding (prefer UTF-8) before modifying files
• Normalize line endings and remove BOMs to avoid parser issues
• Collapse redundant whitespace and trim leading/trailing spaces early
• Replace smart quotes and fancy dashes with ASCII equivalents for compatibility
• Test transformations on a subset and consider edge cases (empty lines, zero-width chars)

Example use cases

• Clean a messy CSV: detect delimiter, normalize encoding, strip BOM, then trim fields
• Prepare text for NLP: remove punctuation, lowercase, collapse spaces, and strip zero-width characters
• Localize content: normalize quotes/dashes, ensure Unicode-aware length handling, and maintain encoding
• Extract contact data: grep emails and URLs from large logs or scraped pages
• Write consistent copy: apply formatting rules and generate headlines or CTAs with templates