we've gone full Inception: building frameworks to detect the frameworks

A hierarchical decision tree/framework for detecting AI-generated content and hybrid human-AI writing. Presented as a technical classification system with 20 distinct linguistic pillars and metrics for identifying LLM signatures. Extracted text: 01_ai_origin_nlp_signals/ (Classifiers 1-20) ├─ definitions/ │ ├─ "hybrid_content": Human + AI-assisted writing with operator-level signal. │ └─ "ai_signature": Detectable statistical uniformity common in LLM outputs. ├─ decision_pillars/ │ ├─ 1) lexical_diversity_index │ ├─ 2) syntactic_burstiness │ ├─ 3) semantic_drift_monitor │ ├─ 4) pattern_repetition_audit │ ├─ 5) pronominal_frequency │ ├─ 6) passive_voice_saturation │ ├─ 7) idiomatic_regionalism │ ├─ 8) transition_word_predictability │ ├─ 9) sentence_complexity_jitter │ ├─ 10) emotional_variance │ ├─ 11) cliche_density │ ├─ 12) rhetorical_question_ratio │ ├─ 13) verb_tense_consistency │ ├─ 14) adverbial_fluff_score │ ├─ 15) proper_noun_density │ ├─ 16) formatting_logic_consistency │ ├─ 17) metaphor_originality │ ├─ 18) nuance_preservation │ ├─ 19) prompt_leakage_detection │ └─ 20) perplexity_score_volatility └─ risk_output: Low | Moderate | High AI Signature