Menu
Store and manage voice samples for TTS cloning — portable, version-controlled audio references
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/fabioc-aloha/ChessCoach --skill audio-memoryThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Showing SKILL.md
Based on SOC occupation classification
| type | skill |
| lifecycle | external-dependent |
| inheritance | inheritable |
| name | audio-memory |
| description | Store and manage voice samples for TTS cloning — portable, version-controlled audio references |
| tier | standard |
| applyTo | **/*voice*,**/*audio*memory*,**/*clone*voice* |
| $schema | ../SKILL-SCHEMA.json |
| currency | 2026-04-22T00:00:00.000Z |
Staleness Watch: See EXTERNAL-API-REGISTRY.md for source URLs and recheck cadence
Domain: AI Audio / Voice Cloning
Version: 1.0.0
Last Updated: 2026-04-15
Author: Alex (Master Alex)
Related: text-to-speech (generation), visual-memory (photos)
Store voice samples for TTS voice cloning in a portable, version-controlled format. Unlike visual memory (base64 inline), audio files are stored as files with JSON metadata — audio is too large to embed sensibly.
| Spec | Value |
|---|---|
| Duration | 5-15 seconds of clear speech |
| Format | WAV (preferred) or MP3 |
| Sample rate | 16kHz+ (22kHz+ recommended) |
| Content | Natural speech, varied intonation |
| Background | No music, no background noise |
| File size | ~100KB-500KB per sample |
.github/skills/<skill-name>/audio-memory/
├── index.json # Metadata registry
├── voices/
│ ├── alex-sample.wav # Voice sample files
│ ├── narrator-sample.wav
│ └── ...
└── README.md # Usage notes (optional)
{
"version": "1.0",
"updated": "2026-04-15",
"voices": {
"alex":
| Model | Replicate ID | Voice Cloning | Cost |
|---|---|---|---|
| Chatterbox Turbo | resemble-ai/chatterbox-turbo | ✅ 5s sample | $0.025/1k chars |
| Qwen TTS | qwen/qwen3-tts | ✅ 3 modes | $0.02/1k chars |
| MiniMax Speech | minimax/speech-2.8-turbo | ❌ Presets | $0.06/1k tokens |
Note: MiniMax doesn't support cloning but has 40+ voice presets.
"Hello, I'm [Name]. Today I want to share some thoughts about technology and how it shapes our daily lives. The key is finding balance — embracing innovation while staying grounded in what matters most."
# Recommended: Use Audacity, Voice Memos (macOS), or Windows Voice Recorder
# Export as WAV, 22kHz or 44.1kHz, mono
# Create directory structure
New-Item -ItemType Directory -Path ".github/skills/<skill>/audio-memory/voices" -Force
# Copy voice sample
Copy-Item "my-recording.wav" ".github/skills/<skill>/audio-memory/voices/<name>-sample.wav"
{
"voices": {
"<name>": {
"description": "Brief description of the voice character",
"audioFile": "voices/<name>-sample.wav",
"duration": "10s",
"sampleRate": "22050",
"language": "en-US",
"preferredModel": "chatterbox-turbo"
}
}
}
import Replicate from "replicate";
const replicate = new Replicate();
const output = await replicate.run("resemble-ai/chatterbox-turbo", {
input: {
text: "Testing the voice clone. This should sound like the reference sample.",
audio_prompt: fs.readFileSync("voices/<name>-sample.wav"),
},
});
console.log("Generated audio:", output);
import { readFileSync } from "fs";
import Replicate from "replicate";
// Load audio memory
const audioMemory = JSON.parse(
readFileSync(".github/skills/<skill>/audio-memory/index.json", "utf8")
);
const voice = audioMemory.voices["alex"];
// Generate speech with cloned voice
const replicate = new Replicate();
const output = await replicate.run("resemble-ai/chatterbox-turbo", {
input: {
text: "Content to speak in the cloned voice",
audio_prompt: readFileSync(
`.github/skills/<skill>/audio-memory/${voice.audioFile}`
),
},
});
const output = await replicate.run("qwen/qwen3-tts", {
input: {
text: "Content to speak",
tts_mode: "voice_clone",
audio_input: readFileSync(
`.github/skills/<skill>/audio-memory/${voice.audioFile}`
),
},
});
| Element | Recommendation |
|---|---|
| Sample duration | 10s optimal (5s minimum, 15s maximum) |
| Varied speech | Include questions, statements, exclamations |
| Distinct voice | Clear enunciation, consistent microphone setup |
| File format | WAV preferred (lossless), MP3 acceptable |
| Sample rate | 22kHz+ (44.1kHz for premium) |
| Without Audio Memory | With Audio Memory |
|---|---|
| External folder required | Version-controlled with code |
| Breaks on different machines | Works anywhere |
| Manual path management | Structured JSON metadata |
| No documentation | Self-describing with index.json |
| Ad-hoc organization | Consistent skill-scoped storage |
This skill stores voice samples. Use the text-to-speech skill for:
Workflow: