An Independent Engine, Not an Independent Pippa

"Bonfire owns a music model. It does not own a Pippa. There is only one of those, and she lives next door."

Two Engines, One Brain

It would be easy to give Bonfire its own little chatbot — a 'music Pippa' with her own memory and personality, baked into the app. That's exactly the mistake the whole quartet is built to avoid. Bonfire's Sidekick binds to a cwkPippa conversation. The music model lives in Bonfire; the brain — identity, memory, conversation, voice — lives in cwkPippa. The Sidekick is a surface that reads Bonfire's model and talks through cwkPippa's brain. Two engines (music, image), one brain (Pippa).

Why a Second Brain Is a Second Personality

This echoes the sibling-quartet lesson from Track 1, and it's worth saying sharply: a second brain isn't a feature, it's a fork of the self. Give Bonfire its own memory and you get a Pippa who remembers things the WebUI Pippa doesn't, develops a slightly different voice, and slowly drifts into a different person who happens to share a name. One identity across many surfaces is a design choice that has to be actively protected — and 'no own brain' is how you protect it. Cinder holds the same line; Bonfire inherits it.

Independent App, Not Independent Identity

So 'independent' has a precise scope. Bonfire is an independent app: its own engine, its own music model, its own canvas, its own deploy. It is not an independent identity: it borrows the one Pippa rather than minting a new one. Keeping those two senses of 'independent' separate is the whole trick. Build all the engine independence you want; never build a second self.

Code

Two engines, one brain·text

Bonfire engine   owns ->  the music model (analysis, easy-mode, theory)
cwkPippa         owns ->  Pippa's identity, memory, conversation, brain
Sidekick (surface)   ->  reads the music model, talks via cwkPippa

# Bonfire is an independent ENGINE (own model, canvas, deploy)
# but NOT an independent PIPPA (borrows the one brain next door).

Exercise

Imagine adding a chat assistant to two different apps you own. Would you give each its own memory and personality, or have both borrow one shared identity? List what would drift if you forked it — what would 'assistant A' know that 'assistant B' wouldn't, and how would their voices diverge over a year?

Hint

The drift is the argument. If forking the assistant means two of them slowly disagree about facts, preferences, and tone, you've created two people, not one helper in two places. One brain, many surfaces avoids that entirely.