Build the voice pack from past scripts.

Step 1: Index scripts into ChromaDB

Run python pipelines/index_voice.py to embed and store all scripts from voice/samples/.

Step 2: Analyze writing style

Read all script files in voice/samples/ and analyze them to extract:

Opening/hook patterns: How do scripts typically start? What makes the hooks work?
Sentence rhythm: Short punchy sentences? Long flowing ones? Mix?
Vocabulary: Casual or formal? Jargon usage? Slang?
Tone: Authoritative? Conversational? Humorous? Provocative?
Structural patterns: How are arguments built? What comes after the hook?
Signature phrases: Words or phrases that appear repeatedly
Transition patterns: How does the creator move between sections?
CTA patterns: How do scripts end? What actions are requested?

Build the voice pack from past scripts.

Run python pipelines/index_voice.py to embed and store all scripts from voice/samples/.

Read all script files in voice/samples/ and analyze them to extract:

Opening/hook patterns: How do scripts typically start? What makes the hooks work?
Sentence rhythm: Short punchy sentences? Long flowing ones? Mix?
Vocabulary: Casual or formal? Jargon usage? Slang?
Tone: Authoritative? Conversational? Humorous? Provocative?
Structural patterns: How are arguments built? What comes after the hook?
Signature phrases: Words or phrases that appear repeatedly
Transition patterns: How does the creator move between sections?
CTA patterns: How do scripts end? What actions are requested?