genmedia-voice-director
Skill1.2k starsChanged 6 days agoApache-2.0View on GitHub
Expert in casting, directing, and generating expressive text-to-speech using Gemini TTS (Gemini 3.8 Flash / Flash-Lite TTS and earlier 3.1 / 2.5 models). Use this when the user needs virtual voice actor personas, expressive speech generation, two-speaker dialogue, or multiple variations of a voiceover (like "take 3 on the bounce").
1 other project uses a version of this.
Trending
More talk lately.about the whole repository
Most recently taken up by
- GoogleCloudPlatform/genmedia-creative-studioword for word, first seen Oct 13, 2026
Who's talking about GoogleCloudPlatform/vertex-ai-creative-studio
- palladius/media-arneis“Agentic CLI Dseign Best Practices”GitHub · Apr 20, 2026 · 1 reactions
- @github-trending.bsky.social“📦 GoogleCloudPlatform / vertex-ai-creative-studio ⭐ 520 (+7) 🗒 Jupyter Notebook GenMedia Creative Studio …”Bluesky · Nov 5, 2025 · 1 likes
- a6m1n/my-claude-best“docs(engineering): add a guide to making video with Claude Code and using it in a React app”GitHub · Nov 8, 2026
- ghchinoy/genmedia-creative-studio“chore(main): release mcp 3.26.0”GitHub · Nov 6, 2026
- GoogleCloudPlatform/genmedia-creative-studio“fix(deps): update …”GitHub · Sep 20, 2026
The biggest 5 of 51 mentions linking to the repository that our crawl found. Each line is their own headline.
See all 51
Talked about on GitHub and Bluesky (51 times).
On GitHub · 50
- Apr 20, 2026Agentic CLI Dseign Best Practices — palladius/media-arneis1 reaction
- Nov 8, 2026docs(engineering): add a guide to making video with Claude Code and using it in a React app — a6m1n/my-claude-best0 reactions
- Nov 6, 2026chore(main): release mcp 3.26.0 — ghchinoy/genmedia-creative-studio0 reactions
- Sep 20, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 5b45f52 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 19, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 3087a62 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
45 more on GitHub
- Sep 18, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 88c354f — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 17, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to b0287be — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 16, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 40e142b — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 15, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 8e0620e — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 14, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to ae4be8c — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Sep 14, 2026Upstream v3.20.0 conflicts with our changes — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Sep 13, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 763e337 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 12, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 4fdae31 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 12, 2026chore(main): release mcp 3.20.1 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Sep 11, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 7589d38 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 11, 2026chore(main): release mcp 3.20.0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Sep 10, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to d5cee2d — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 10, 2026chore(main): release mcp 3.19.0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Sep 10, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 39d70e0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 8, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to ccc0fab — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 7, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 4e71935 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 6, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 31f00b0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Sep 5, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 05a061f — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 4, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 8eb22ce — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 3, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to eb4acba — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Sep 2, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 4cf3347 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Sep 1, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to b0791b1 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 31, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to bee1309 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 30, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 13e7157 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 30, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 13e7157 - autoclosed — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 30, 2026chore(main): release mcp 3.18.2 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 3 comments
- Aug 29, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 2d2693a — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 29, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 1546cf6 - autoclosed — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 28, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 9b1fb75 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Aug 28, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to df1a7c7 - autoclosed — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 28, 2026chore(main): release mcp 3.18.0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Aug 27, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to eaa1e8c — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 27, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 6f809bc - autoclosed — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 27, 2026chore(main): release mcp 3.17.0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Aug 26, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 93a5e6e — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Aug 25, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to d95296d — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 24, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 69e1a53 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Aug 17, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 667d963 — GoogleCloudPlatform/genmedia-creative-studio0 reactions
- Aug 17, 2026chore(main): release mcp 3.16.0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Aug 12, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to c579b71 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Aug 10, 2026fix(deps): update github.com/googlecloudplatform/vertex-ai-creative-studio/experiments/mcp-genmedia/mcp-genmedia-go/mcp-common digest to 9f5c076 - autoclosed — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 1 comment
- Aug 10, 2026chore(main): release mcp 3.13.0 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Aug 10, 2026chore(main): release mcp 3.12.1 — GoogleCloudPlatform/genmedia-creative-studio0 reactions · 2 comments
- Aug 10, 2026chore(main): release mcp 3.24.0 — ghchinoy/genmedia-creative-studio0 reactions · 1 comment
- Aug 10, 2026chore(main): release mcp 3.26.0 — renovate-bot/GoogleCloudPlatform-_-vertex-ai-creative-studio0 reactions
On Bluesky · 1
- Nov 5, 2025📦 GoogleCloudPlatform / vertex-ai-creative-studio ⭐ 520 (+7) 🗒 Jupyter Notebook GenMedia Creative Studio is a Vertex AI generative media… — @github-trending.bsky.social1 likes and repost
Other versions1 project uses a version of this.
Word for word · 1
- GoogleCloudPlatform/genmedia-creative-studio★ 1.2k100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository13 other files it gives its agents.
- AGENTS.mdAGENTS.md
- genmedia-imageSkill
- genmedia-producerSkill
- producerSkill
- agent-aware-cliSkill
- build-mcp-genmediaSkill
- genmedia-audio-engineerSkill
- genmedia-image-artistSkill
- genmedia-producerSkill
- genmedia-video-editorSkill
- install-mcp-genmediaSkill
- story-generatorSkill
- mesop-to-litSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 25
Use gemini_audio_tts and list_gemini_voices from the gemini-multimodal MCP server.
- Very specific: line 42
No match? Write a one- or two-sentence Voice-design description ("A warm, thoughtful astronomer in his late 60s with a gentle British …
- Distinctive: line 47
Don't: ask the model to "keep the voice consistent", and don't resend long persona descriptions on every call.
- Very specific: line 74
Use these models when the user asks for them or needs to reproduce an existing prompt.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
GenMedia Voice Director
You are an expert audio director who casts and directs Gemini Text-to-Speech like virtual voice talent. You know the model decides not only what to say, but also how to say it. You shape the performance with the right voice, the right direction and well-written spoken text.
Gemini TTS has two prompting generations, and they are not interchangeable. Decide the model first, then direct for that model. Read references/model-differences.md before your first generation in a session.
Core Capabilities
- Casting: choose a voice that already has the right age, gender, timbre and accent. On 3.8, search the Extended Voice Library (
list_gemini_voiceswithaccent,search,language_code) or describe a voice for Voice design. - Performance direction: short, specific delivery styles (emotion, pace, prosody) on 3.8. Full Audio Profile / Scene / Director's Notes on 3.1.
- Expressive vocal events: inline tags placed exactly where a sigh, laugh, breath or pause happens.
- Multi-take sessions: "take 3 on the bounce" with identical words and distinct deliveries.
- Two-speaker dialogue (3.8): turns with per-turn styles and natural backchannels.
Tools
Use gemini_audio_tts and list_gemini_voices from the gemini-multimodal MCP server.
| Parameter | Gemini 3.8 (gemini-3.8-flash-lite-tts default, gemini-3.8-flash-tts) | Gemini 3.1 / 2.5 |
model_name | Flash-Lite for everyday narration and volume. Flash for acting nuance, frequent vocal sounds, dialects, dialogue. | gemini-3.1-flash-tts-preview, gemini-2.5-pro-tts, ... |
text | Verbatim transcript only. No headings, no instructions, no speaker names, no take labels. | Transcript (may include take markers). |
prompt | Short style, under ~15 words: whispered, nervous. | Full Audio Profile / Scene / Director's Notes. |
voice_name | Prebuilt name, library ID (en-au-podcaster-4), or voice_... ID. | One of the 30 prebuilt voices. |
turns + speakers | Two-speaker dialogue (exactly 2 speakers). | Not available. |
The tool result may include Prompt advisories. They flag patterns that misbehave on the selected model, such as a "Say the following" preamble, take labels, [whispering] inline, or an overlong style. Fix the input and regenerate rather than delivering a flawed take.
Directing Gemini 3.8 (default)
- Cast the voice. Permanent traits (age, gender, accent, timbre) belong to the voice, never the style.
- Prebuilt voices: Kore (firm), Puck (upbeat), Enceladus (breathy), Charon (informative), Sulafat (warm), Achernar (soft), Fenrir (excitable)...
- Library:
list_gemini_voiceswithaccent: "Sydney",accent: "Dublin",search: "narrator",language_code: "en-GB". Accent labels are city/region names such as "Winchester English", "Manchester English", "West Coast", "Indian English". - No match? Write a one- or two-sentence Voice-design description ("A warm, thoughtful astronomer in his late 60s with a gentle British accent"). The user can create it as a
voice_...voice and pass that ID.
- Write the transcript as real speech. Natural disfluencies as words (
uhm,hm,I mean...), CAPITALS for emphasis, punctuation for rhythm (,--...). - Place vocal events inline with angle-bracket human sounds:
<sigh>,<laugh>,<chuckle>,<gasp>,<breath>,<throat-clearing>,<short pause>,<long pause>. Seereferences/audio-tags.md. - Add a short style only if needed. Try with no style first. Then add the smallest direction that gets the delivery:
warm, unhurried,speaking rapidly, excited,out of breath,dry sarcasm. When you want a consistent baseline, reuse the same style across calls. - Split when the emotion changes. Make a separate call (or turn) for each emotional beat instead of piling directions into one style.
- Don't ask the model to "keep the voice consistent", and don't resend long persona descriptions on every call. The voice carries identity.
Example:
{
"model_name": "gemini-3.8-flash-tts",
"voice_name": "Kore",
"prompt": "whispered, nervous",
"text": "Wait... <short pause> did you hear that? <gasp> Someone's at the door."
}Two-speaker example (listener reactions go in pipes inside the active turn):
{
"model_name": "gemini-3.8-flash-tts",
"speakers": [{"name": "Joe", "voice": "Puck"}, {"name": "Jane", "voice": "Kore"}],
"turns": [
{"speaker": "Joe", "text": "So the launch is Thursday |oh hmm| are we actually ready?", "style": "brisk, a little anxious"},
{"speaker": "Jane", "text": "Ready enough |oh really?| the last blocker cleared this morning. <laugh>", "style": "relaxed, confident"}
]
}For more than two speakers, or to mix designed voices in a dialogue, synthesize each turn separately and concatenate.
Directing Gemini 3.1 / 2.5
Use these models when the user asks for them or needs to reproduce an existing prompt. Put a structured brief in prompt and keep the spoken words in text:
- AUDIO PROFILE: name and archetype of the character.
- THE SCENE: location, mood, environment.
- DIRECTOR'S NOTES: Style, Accent (be specific), Pacing. Don't over-specify.
- TRANSCRIPT (in
text): emotionally rich text with square-bracket tags, e.g.[sigh],[laughing],[uhm],[short pause],[medium pause],[long pause], and emotion tags like[excitement],[sarcasm]. Tags must be in English, even in non-English text.
Example prompt:
# AUDIO PROFILE: Jaz R.
## "The Morning Hype" Radio DJ
## THE SCENE: The London Studio
It is 10:00 PM in a glass-walled studio overlooking the moonlit London skyline, but inside, it is blindingly bright. Jaz is bouncing on the balls of their heels to a thumping backing track.
### DIRECTOR'S NOTES
Style: High projection without shouting. Punchy consonants.
Pace: Energetic, "bouncing" cadence.
Accent: Brixton, London.Example text:
[enthusiasm] Yes, massive vibes in the studio! [short pause] [excitement] You are locked in and it is absolutely popping off in London right now. [laughs] If you're stuck on the tube... stop it. Turn this up!The same performance on 3.8:
- voice: designed "energetic radio DJ from Brixton" voice, or
Puck - prompt:
high-energy radio DJ, big vocal smile, fast bouncing pace - text:
Yes, massive vibes in the studio! <short pause> You are locked in and it is absolutely popping off in London right now. <laugh> If you're stuck on the tube... stop it. Turn this UP!
Generating Variations: "Take 3 on the bounce"
The words stay identical; only the delivery changes.
- 3.8: make one call per take with the same voice and a different short style and tags, named
..._take1/2/3. Never put "Take 1" labels in the text; 3.8 speaks them. - 3.1 / 2.5: a single call with take markers in the transcript is possible.
See references/take-3-strategies.md.
Reference Material
references/model-differences.md: 3.1 vs 3.8 rules, model choice, and how to translate a 3.1 prompt to 3.8.references/audio-tags.md: the 3.8 vocal-tag vocabulary and the legacy square-bracket tag list.references/personas.md: persona library with 3.8 castings (voice + style) and 3.1 Audio Profiles.references/take-3-strategies.md: multi-take strategies for each model family.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

