Docs Engineer — Bundle

The problem

Stale examples, bloated explanations, and docs organized by code structure rather than user intent create friction for developers. Every undocumented prerequisite or unverified code snippet leads to broken builds and wasted support time.

What it does

• Identifies the reader's immediate task and prerequisites before drafting content.
• Verifies command-line instructions and code examples against your actual project files.
• Flags stale or redundant documentation for deletion to prevent knowledge rot.
• Refactors long-form descriptions into concise, runnable worked examples.
• Enforces technical accuracy by marking any unverified claims as untrusted.

Why this beats prompting it yourself

Generic LLMs tend to produce "comprehensive" but fluffy documentation that users won't read. This skill enforces a ruthlessly scoped, reader-first engineering mindset that treats documentation as a tested product rather than a writing exercise.

Use cases

• Drafting task-oriented Getting Started guides from a fresh codebase.
• Auditing existing READMEs for stale commands and broken code snippets.
• Converting complex API technical specs into concise developer tutorials.
• Standardizing documentation tone across a multi-repo engineering team.

Known limitations

Does not author marketing copy, legal text, or licenses. Cannot make architectural decisions or redesign your API; it only documents existing systems.