Claude Code Plugins

Community-maintained marketplace

Feedback

say-narration

@plurigrid/asi
0
0

Use macOS text-to-speech for agent narration and announcements. Sub-agents announce themselves using different language voices speaking English. Use for multi-agent workflows where each agent has a distinct voice identity.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name say-narration
description Use macOS text-to-speech for agent narration and announcements. Sub-agents announce themselves using different language voices speaking English. Use for multi-agent workflows where each agent has a distinct voice identity.
version 1.0.0

Say Narration (macOS TTS)

Use macOS text-to-speech for agent announcements and narration.

MANDATORY: NextColor Per Interaction

EVERY interaction MUST use NextColor with the voice name as the seed.

;; Before EVERY voice announcement, compute NextColor
(def color (next-color (seed-from-string "Ava (Premium)")))
;; color determines the trit assignment: MINUS(-1), ERGODIC(0), PLUS(+1)

The voice name string IS the seed for deterministic color assignment. This ensures GF(3) conservation across all voice interactions.

Quality Requirements

ONLY use Enhanced or Premium quality voices. NEVER use:

  • Base/standard quality voices (no suffix)
  • British man voice (Daniel)
  • Any novelty voices (Albert, Bad News, Bells, Boing, etc.)

Approved High-Quality Voices

bmorphism Mathematician Personas (Premium)

Voice Language Mathematician Persona Haiku Theme
Anna (Premium) German Emmy Noether Symmetry, Algebra
Emma (Premium) Italian Maria Adelaide Sneider Algorithms dance
Federica (Premium) Italian Pia Nalli Theorems flow
Serena (Premium) English UK Bertha Swirles Quantum waves
Petra (Premium) German Ruth Moufang Algebra speaks
Yuna (Premium) Korean Hee Oh Hidden patterns
Alva (Premium) Swedish Sonja Korovkin Patterns flow
Amélie (Premium) French CA Sophie Germain Prime numbers
Ewa (Premium) Polish Maria Wielgus Logic roots
Kiyara (Premium) Hindi Shakuntala Devi Numbers dance
Majed (Premium) Arabic Maha Al-Aswad Numbers dance
Tünde (Premium) Hungarian Julia Erdős Numbers soar
Han (Premium) Chinese Chen Jingrun Prime dancing
Lilian (Premium) Chinese Hua Luogeng Number theory
Sinji (Premium) Chinese HK Shing-Tung Yau Manifolds reveal
Yue (Premium) Chinese Chern Shiing-shen Differential forms

Currently Installed Voices (221 total, 14 Premium/Enhanced)

Voice Quality Language Persona Seed
Ava (Premium) Premium en_US - "Ava (Premium)"
Anna Standard de_DE Emmy Noether "Anna (German (Germany))"
Amélie Standard fr_CA Sophie Germain "Amélie (French (Canada))"
Milena Standard ru_RU Olga Ladyzhenskaya "Milena (Russian (Russia))"
Tingting Standard zh_CN Wang Zhenyi "Tingting"
Sinji Standard zh_HK Shing-Tung Yau "Sinji"
Ava (Enhanced) Enhanced en_US - "Ava (Enhanced)"
Allison (Enhanced) Enhanced en_US - "Allison (Enhanced)"
Samantha (Enhanced) Enhanced en_US - "Samantha (Enhanced)"
Nathan (Enhanced) Enhanced en_US - "Nathan (Enhanced)"
Evan (Enhanced) Enhanced en_US - "Evan (Enhanced)"
Nicky (Enhanced) Enhanced en_US - "Nicky (Enhanced)"
Noelle (Enhanced) Enhanced en_US - "Noelle (Enhanced)"
Alice (Enhanced) Enhanced it_IT - "Alice (Enhanced)"
Emma (Enhanced) Enhanced it_IT Maria Sneider "Emma (Enhanced)"
Federica (Enhanced) Enhanced it_IT Pia Nalli "Federica (Enhanced)"
Paola (Enhanced) Enhanced it_IT - "Paola (Enhanced)"

Install Missing Voices

Open System Settings > Accessibility > Spoken Content > System Voice > Manage Voices

Priority installs for GF(3) triadic coverage:

  1. Anna (Premium) - German - MINUS validator voice
  2. Amélie (Premium) - French - ERGODIC coordinator voice
  3. Yuna (Premium) - Korean - PLUS generator voice

Usage Pattern (MANDATORY)

# Step 1: Compute NextColor with voice name as seed
# (conceptually - the color determines agent role)

# Step 2: Announce with high-quality voice
say -v "Ava (Premium)" "Agent Oracle reporting status"
say -v "Allison (Enhanced)" "Technical analysis complete"
say -v "Nathan (Enhanced)" "Macro calculations finished"

Agent Voice Assignments (Updated)

Agent Voice Trit Role
Agent-Nash-Oracle Ava (Premium) NextColor("Ava (Premium)")
Agent-Technical-Vulture Nathan (Enhanced) NextColor("Nathan (Enhanced)")
Agent-Macro-Basalt Allison (Enhanced) NextColor("Allison (Enhanced)")
Agent-Exa-Stalker Alice (Enhanced) NextColor("Alice (Enhanced)")
Agent-Contrarian-Phoenix Evan (Enhanced) NextColor("Evan (Enhanced)")
Agent-Temporal-Specter Samantha (Enhanced) NextColor("Samantha (Enhanced)")
Agent-Theta-Harvester Emma (Enhanced) NextColor("Emma (Enhanced)")
Agent-Narrative-Proteus Noelle (Enhanced) NextColor("Noelle (Enhanced)")

List Available High-Quality Voices

say -v '?' | grep -E "(Enhanced|Premium)"

GF(3) Conservation

Each voice interaction contributes to the triadic balance:

  • Seed = voice name string
  • Color = NextColor(seed) → {-1, 0, +1}
  • Sum across session must satisfy Σ trits ≡ 0 (mod 3)

Tips

  • Italian Enhanced voices speaking English creates distinctive character
  • Use Premium voices for primary agents (best quality)
  • Never use Daniel or any British male voices
  • Never use novelty/effect voices

Scientific Skill Interleaving

This skill connects to the K-Dense-AI/claude-scientific-skills ecosystem:

Graph Theory

  • networkx [○] via bicomodule
    • Universal graph hub

Bibliography References

  • general: 734 citations in bib.duckdb

Cat# Integration

This skill maps to Cat# = Comod(P) as a bicomodule in the equipment structure:

Trit: 0 (ERGODIC)
Home: Prof
Poly Op: ⊗
Kan Role: Adj
Color: #26D826

GF(3) Naturality

The skill participates in triads satisfying:

(-1) + (0) + (+1) ≡ 0 (mod 3)

This ensures compositional coherence in the Cat# equipment structure.