What problem does it solve? Creators often guess at account positioning from labels, trends, or wishful thinking, ending up with positioning that cannot survive contact with platform reality. This Skill turns vague ideas into a staged, evidence-backed account positioning by separating internal facts, platform evidence, and publishing feedback, and by refusing to fabricate conclusions when evidence is missing. ## Core Features & Use Cases - Fast or Deep Positioning Modes: Offers a quick path to an executable working positioning (L2) or a deep path that resolves candidate conflicts, content roles, and optional homepage profile (avatar, nickname, bio). - Evidence-Gated Outputs: Without qualified platform evidence pasted back by the user, it only delivers personal material directions plus one custom search task, never fake positioning candidates. - Topic-to-Positioning Bridge: When a user asks whether candidate topics fit the current account stage, it maps each topic to one functional role (main line, persona proof, traffic entry, commercial bridge, or auxiliary) and designs one falsifiable publishing experiment. - Use Case: A creator says "help me do account positioning." The Skill asks once whether they want fast or deep positioning, extracts personal material directions from their real history, assigns one tailored platform search task, and only after the user pastes back evidence forms a staged positioning candidate. ## Quick Start Ask the assistant to run an account staged positioning with eva-positioning and choose either fast positioning or deep positioning when prompted.
