human-writing-azzar_skill

Overview

This skill defines guidelines and standards for producing professional, human-like documentation and technical writing. It targets prose used in READMEs, technical docs, code comments, and PR descriptions, removing common AI stylistic faults. The goal is concise, information-dense output that reads like a senior practitioner wrote it.

How this skill works

When invoked, the skill scans a draft for known LLM 'tells' and banned buzzwords, then restructures sentences to favor precision and verifiable details. It enforces a checklist: remove fluff, substitute technical terms for marketing language, and validate versioning, numbers, and references. The final pass audits parallel constructions and removes emojis or corporate filler.

When to use it

• Writing or revising README files for libraries, tools, or services
• Creating API docs, design docs, or architecture overviews
• Composing code comments, inline explanations, and docstrings
• Preparing PR descriptions, release notes, or changelogs
• Converting high-level marketing text into technical documentation

Best practices

• Prefer concrete facts: versions, constraints, examples, benchmarks
• Avoid buzzwords; use precise technical terms instead
• Keep sentences short and focused; aim for one idea per sentence
• Include references or RFCs when claiming standards or behaviors
• Run a final self-audit for common AI patterns and remove them

Example use cases

• Turn a draft README into a compact, versioned quickstart with clear commands
• Refine API documentation to include parameter types, error codes, and examples
• Edit code comments to explain intent and edge cases without promotional language
• Rewrite release notes to list concrete changes, migration steps, and compatibility
• Convert user-facing marketing copy into technical onboarding steps