What problem does it solve?
AI-generated or AI-assisted fiction often reads mechanically: repeated explanations, overly uniform sentence rhythms, characters who all speak alike, abstract judgments without concrete action, and decorative filler sentences. This Skill diagnoses those patterns and applies minimal, voice-preserving revisions instead of blanket rewrites.
Core Features & Use Cases
- Pattern Diagnosis: Scans text for repeated explanations, uniform sentence structures, homogeneous dialogue, abstract judgments, generic body reactions, empty atmosphere, and functionless aphorisms using the method in references/method.md.
- Minimal Revision: Fixes issues with the smallest possible edits—deleting redundant explanation, converting abstract judgments into consequential actions, and giving dialogue purpose and subtext—without altering facts, viewpoint, character motivation, or confirmed voice.
- Graded Reporting: Ranks issues by reading impact, outputs statistics and before/after samples, and only produces a full polished draft when explicitly requested.
- Use Case: A web novel author pastes a chapter draft that feels robotic; the Skill reports the dominant mechanical patterns with locations and suggested revisions, then produces a refined version only after confirmation.
Quick Start
Use oh-story-deslop to scan this chapter for AI-sounding patterns and suggest minimal revisions that keep my characters' voices intact.
