Score articles across seven dimensions and flag factual risks.
npx skills add https://github.com/hayatosc/dotfiles --skill article-review-hayatosc
Please help me install this Agent Skill. Skill: article-review Source: https://github.com/hayatosc/dotfiles/tree/main/home/dot_agents/skills/article-review Command: npx skills add https://github.com/hayatosc/dotfiles --skill article-review-hayatosc
Many writers and teams need a short, rigorous editorial pass that assesses publishability, uncovers factual or social risks, and gives prioritized, high-leverage fixes plus example rewrites without replacing the author's voice. This Skill fills the gap between casual feedback and a full human edit by producing structured scores, safety checks, and concrete rewrite examples that editors can act on quickly.
Review this article and return the chosen mode, seven-dimension scores, a safety assessment, the top three prioritized fixes, and up to three concise rewrite examples.
High-intent search queries and answers about installing and using this skill.
Editorial review evaluates written drafts by assigning seven-dimension scores, flagging factual and social risks, and proposing prioritized fixes with targeted rewrites to ensure publication safety and clarity.
Fact-checking during editorial review involves enabling the verifyFacts flag to surface claims needing verification and identify social or reputational risks with specific remediation suggestions.
Yes, editorial review provides one to three targeted rewrite examples that demonstrate prioritized fixes while deliberately preserving the original author's voice and expression.
Yes, editorial review supports both modes by classifying the piece as information-first or empathy-first, noting mixed purposes, and applying a strict scoring rubric tailored to each mode.
Scoring an article involves applying a strict 0.0–5.0 reference rubric across seven dimensions—reliability, structure, reader value, expression, tone, safety, and human distinctiveness.
Automated editorial review bridges the gap between casual feedback and full human edits by providing structured scores and rewrites, but requires full text input and optional context flags to function effectively.
groundnutyExtract factual claims from text and verify them via a forked claim-verifier.
hyeshikExtract key insights from scholarly papers into structured interview drafts.
synthesisengineeringEvaluate AI-assisted content quality using a 36-criteria checklist.
MuvonDetect AI-writing signatures and rewrite drafts with an eight-dimension scoring and seven-pass process.
