What problem does it solve? Drafting and reviewing content in a specific owner's voice produces inconsistent results when the writer and reviewer apply different mental rules. This Skill gives both personas one shared, ratified calibration — the goal, the filter, the subject bound, corpus evidence, register constraints, and teaser arc — so a review produces a conflict against the same ruler rather than two unrelated opinions. ## Core Features & Use Cases - Voice calibration rules: Encodes the owner's stated directives (proximity and informality, no authority positioning, touching the reader's personal side) with precedence rules resolving conflicts between the platform architecture page, the owner's live interview statements, and his 26-article Medium corpus. - Register as mechanism: Defines the 'poetry for developers' register as four checkable constraints — claims resolving to openable files, admissions that cost the author, limits stated first, and short declaratives after long technical passages — rather than an unreproducible style. - Review-ready findings: Gives a content-reviewer persona ranked, citable rules (including numbered downstream rules) to raise findings against drafts of articles, site copy, and LinkedIn/X teasers before they reach the owner. - Use Case: A drafting persona writes a diary-style article about the owner's agent loop; a reviewing persona checks it against the same file and flags a passage that positions the author instead of delivering value, citing the exact rule it violates. ## Quick Start Review this draft article against the published-voice calibration and list every passage that violates a ranked rule.
