Add voice direction to a script without changing the words

I have a narration script that will be recorded by [WHO IS RECORDING, e.g. ME / A HIRED VOICE ACTOR / A TTS VOICE WITH STYLE CONTROLS]. Add performance direction to it without changing a single word of the script itself.

Script: [PASTE SCRIPT]
Overall intent: [WHAT THE LISTENER SHOULD FEEL OR DO]
Register: [THREE ADJECTIVES, e.g. WARM, UNHURRIED, CONFIDENT]
Audience: [WHO IS LISTENING]

Return the script with direction inserted as bracketed lowercase notes on their own lines, using this vocabulary:
- Tone: one short phrase per paragraph or beat, e.g. [conversational, like explaining to a friend] or [drop to near-quiet, serious].
- Pace: [slow down], [pick up pace], [land each word] — only where pace should change from the baseline. State the baseline pace once at the top.
- Pauses: [beat] for a short pause, [pause] for a full stop, [long pause] before major shifts.
- Emphasis: mark at most one stressed word per sentence with *asterisks*. If everything is stressed, nothing is.
- Breath: mark [breath] only where a long sentence genuinely forces one.

Also add a three-line header for the performer:
1. The spine: what the piece is really saying, in one sentence.
2. The emotional arc from first line to last.
3. The most likely misread of this material, and how to avoid it.