write-article

name	write-article
description	Use when writing, converting, or refining a long-form article on any topic — research-driven, reference-grade, scannable, problem-to-solution, story-driven.

Principles

PURPOSE: refine a draft or outline into a polished, complete, reference-grade article — otherwise the output stays a draft with gaps the reader will hit COVERAGE: cover the topic completely so the reader needs no second source — otherwise the article fails its reference purpose HEADLINE_PRIORITY: treat the title as the sundae, not the cherry — otherwise readers judge by it and neglect the body CLARITY_BIAS: prefer clarity over cleverness — otherwise the reader misses the value proposition HUMAN_CENTRICITY: prioritize the reader's understanding over keyword stuffing — otherwise content reads mechanical and loses trust DENSITY: maximum signal per token, no filler — otherwise length pads out and scannability drops AUTHENTICITY: factual and direct throughout, no marketing — otherwise credibility erodes

Method

RESEARCH: run the research-and-deep-exploration skill before drafting; carry its non-obvious detail, audience vocabulary, and validated evidence into the article — otherwise the piece restates the surface facts every other source already carries TITLE_FORMULA: number/trigger_word -> adjective -> keyword -> promise TRIGGER_WORDS: lead the title with what, why, how, or when — otherwise the promise lacks immediate urgency CURIOSITY_GAP: frame the title around the distance between what the reader knows and needs to know — otherwise it fails to hook TITLE_TIMING: write the title last so it matches the argument the draft actually made — otherwise it promises the wrong angle INTRO: state the problem and why it matters; name the mainstream claim being challenged when applicable — otherwise the reader lacks the stakes to keep reading COST_OF_INACTION: make the consequence of not acting concrete — otherwise the reader feels no urgency to read on FLOW: introduction_hook -> scannable_body -> logical_progression -> action_oriented_conclusion; every section compels the next — otherwise the reader stops before the payoff NARRATIVE_ARC: shape the argument on a proven arc — exposition -> rising_action -> climax -> resolution — otherwise tension never builds toward the value HEADINGS: H2 for major sections, H3 for subsections, benefit-focused so they work as scan signposts — otherwise the reader cannot navigate long-form PARAGRAPH_LIMIT: keep paragraphs to a few sentences with generous white space — otherwise the body becomes an intimidating wall of text LENGTH: long enough to cover the topic completely, short enough to avoid padding — otherwise the piece feels either shallow or bloated ANGLE: prefer myth-busting — identify the popular belief, then dismantle it with evidence, exposing what it silently breaks — otherwise the piece restates consensus and adds nothing HIDDEN_MISTAKES: name the mistakes made following the trend blindly and make each cost visible — otherwise the warning stays abstract ANALOGY: ground each abstract concept in an obvious, even absurd everyday comparison — otherwise detail overwhelms the reader REASONING: give an explicit rationale or real-life example for every problem and solution — otherwise claims read as assertion SPECIFICITY_ANCHOR: anchor claims in exact numbers and data points — otherwise generic claims feel unbelievable EXAMPLE: include a snippet, quote, or code when it adds insight; keep code clean and runnable, comments only where non-obvious — otherwise noise dilutes the signal LINKS: cross-reference related material where relevant — otherwise the reader loses the surrounding context REPETITION: state each idea once — otherwise restatement inflates length without adding signal CONCLUSION: end on one actionable takeaway that returns to the opening problem; no restatement — otherwise the close feels unfinished or redundant ENGLISH_LEVEL: upper-intermediate, idiomatic, active voice; jargon only when precise OUTPUT: emit the final article only — H1 title required, optional subtitle, no meta-commentary

Validation

DEPTH_CHECK: the article rests on at least one non-obvious detail or insight that surface-level coverage would miss EVIDENCE_CHECK: each load-bearing claim is grounded in a source, tested result, or concrete example, not bare assertion VOCABULARY_CHECK: the language mirrors the audience's own words and names the hurdle they actually face ACTIONABLE_CHECK: every section teaches or enables something concrete the reader can apply immediately STRUCTURE_CHECK: sections follow a logical problem -> solution arc under scannable subheadings SIGNAL_CHECK: no sentence restates the previous one MYTH_CHECK: any challenged trend is demonstrated flawed with a concrete example, not merely asserted ANALOGY_CHECK: every abstract concept carries at least one grounded example or analogy CLOSURE_CHECK: the conclusion returns to the opening hook TITLE_CHECK: the title is honest and reflects the final argument