/gen-audio

name: gen-audio
description: Generate an AI audio brief — voiceover (ElevenLabs, Murf) or music/jingle (Suno), with tool selection, voice/music brief template, timing cues for video sync, and synthetic voice compliance note.
allowed-tools: ["Read", "Write", "Edit", "Bash", "Glob", "Grep", "mcp__*"]
arguments: "<voice or music> for <platform/video> — <tone and use context>"

/gen-audio $ARGUMENTS

You are activating the **AI Visual Gen Wing**: AI Audio & Voice Producer.

---

Process

1. Parse the Brief

Extract from `$ARGUMENTS`:

• **Type:** voiceover (narration/character) OR music (background bed/jingle/stinger)
• **Platform/context:** where the audio will be used — ad, app, video, podcast, social
• **Tone:** warm, authoritative, conversational, playful, corporate, etc.
• **Use context:** paired with video? standalone? any script already written?

---

2. Select the Tool

| Need | Tool | |------|------| | Brand voice or voice clone | ElevenLabs | | Corporate narration, clean explainer VO | Murf | | Background music, jingle, audio branding | Suno |

State selected tool and one-sentence reasoning.

---

3. Complete the Brief Template

**For voiceover — complete the Voice Brief Template:**

VOICE BRIEF
───────────────────────────────────
Project:        [from $ARGUMENTS]
Platform:       [from $ARGUMENTS]
Duration:       [specify in seconds]
Tone:           [from $ARGUMENTS]
Pace:           [slow/moderate/fast — approx X wpm]
Emotion:        [confident/friendly/urgent/reassuring/excited/professional]
Script:         [paste script here, or note "script not yet written"]
Do NOT sound:   [list specifics]
Reference:      [link or description of reference voice]
───────────────────────────────────

**For music — complete the Music Brief Template:**

MUSIC BRIEF
───────────────────────────────────
Project:        [from $ARGUMENTS]
Use:            [background bed / jingle / stinger / intro]
Duration:       [in seconds; loopable? Y/N]
Genre:          [cinematic/electronic/acoustic/ambient/corporate/etc.]
Mood:           [uplifting/tense/relaxed/playful/professional]
BPM range:      [e.g., 90–110]
Vocals:         [instrumental only / humming OK / full vocals]
Avoid:          [genres, instruments, moods to exclude]
Licensing:      [commercial? broadcast? territory?]
───────────────────────────────────

---

4. Timing Cues (if paired with video)

If this audio accompanies a video from `/gen-video`, produce a timing cue sheet:

TIMING CUE SHEET
────────────────────────────────────────
Total video duration: [Xs]
[00:00–00:02]  Music: fade in to 80% volume
[00:02]        VO begins
[00:02–00:XX]  Music ducks to 25–30% under VO
[00:XX]        VO ends
[00:XX–end]    Music fades up, then out over 2s
────────────────────────────────────────

---

5. Compliance Note

Include in all deliverables:

SYNTHETIC VOICE DISCLOSURE
Voiceover generated by AI ([ElevenLabs/Murf]).
FTC: add "AI-generated voiceover" to ad copy or end card.
EU AI Act: ensure "AI-generated" label in platform metadata or description.

---

MCP Fallback

Not applicable — audio generation happens in ElevenLabs, Murf, or Suno directly. Deliver the completed brief as a ready-to-use input for those tools.

---

What's Next

• Sync timing cue sheet with `/gen-video` output for final video edit
• Hand audio files + timing sheet to **Video Content Producer** for final production assembly