brand-voicelisted

# Brand Voice Tuner Maintains one file: `/media/data/Dropbox/Work/Content/Brand Voice/brand-voice.md` — the canonical source for Mark's voice, broken into a universal **Core voice** and a set of **Contexts** (situational registers: what changes, what stays fixed). This is Mark's personal/general voice reference — broader than the studio-only agent-facing summary in `context/internal/brand.md` (which now just points here). --- ## Step 1 — Determine mode - **No argument**: show the list of contexts already documented (section headers under `## Contexts`) and ask which one to tune, or whether to run `audit`. - **"show" / "list"**: read and print the full master doc, then stop. - **"audit"**: go to Step 5 (consistency check across contexts). - **Anything else** (a context name or description, e.g. "cold outreach emails", "LinkedIn comments", "invoice chase-up emails"): go to Step 2. If the master doc doesn't exist yet, create it first (Step 1a) before continuing. ### Step 1a — First-run bootstrap If `/media/data/Dropbox/Work/Content/Brand Voice/brand-voice.md` doesn't exist, create it with: - A `## Core voice` section seeded from `context/internal/brand.md`'s existing Do/Don't lists (the rules that apply everywhere, regardless of context) - An empty `## Contexts` section with one pre-populated example ported from `context/internal/brand.md`'s "Client messages" section, since that's already a real context-specific register Mark defined - A short header explaining