SKILL: Character Age Drift

Version: 1.0 | Hermes Agent Failure Pattern Library

DESCRIPTION

Mastery of preventing and correcting age inconsistency in AI-generated characters across multiple images or video frames. Age drift is a form of identity degradation where a character's apparent age shifts — younger or older — between generations despite identical age specifications. This is particularly critical in video workflows where temporal coherence requires stable biological age [^87^].

TRIGGER KEYWORDS

age drift, character aging, age inconsistency, age change, looking younger, looking older, age continuity, youth creep, age lock, character age, consistent age, age stability, aging character, skin age

CORE RULES

• Vague age descriptors cause drift: "young adult" is interpreted differently each time
• Anchor age with specific decade + anatomical detail: "35-year-old, slight smile lines, firm jawline"
• Lighting changes cause apparent age shift: soft light reads younger, harsh light reads older
• Include anatomical age anchors: "visible nasolabial folds", "smooth forehead", "no gray hair"
• Expression affects perceived age: smile creates fuller cheeks (younger); serious = hollow cheeks (older)
• For series: lock age description in the identity clause that repeats across all prompts
• Specify explicitly: "age-stable across all frames" in video sequence prompts