prompt: Type.Optional(Type.String({ description: "Music generation prompt." })),

lyrics: Type.Optional(

Type.String({

description: "Optional lyrics to guide sung output when the provider supports it.",

description:

"Exact sung lyrics only when the user supplies lyrics or asks for vocal words. For song/style requests, use prompt instead.",

}),

),

instrumental: Type.Optional(

name: "music_generate",

displaySummary: "Generate music",

description:

'Generate music using configured providers. In session-backed chats, generation runs as a background task; do not call music_generate again for the same request, wait for the completion event, then send the generated attachments through the message tool. Use action="status" to inspect the active task.',

'Create actual audio/music tracks from song, jingle, beat, loop, soundtrack, anthem, or instrumental requests. If the user asks to make/generate/create a song or music, call music_generate; do not just write lyrics unless they ask for lyrics/text only. Put style, genre, mood, tempo, instruments, and purpose in prompt. Use lyrics only for exact sung words. In session-backed chats, generation runs as a background task; do not call music_generate again for the same request, wait for the completion event, then send the generated attachments through the message tool. Use action="status" to inspect the active task.',

parameters: MusicGenerateToolSchema,

execute: async (_toolCallId, rawArgs) => {

const args = rawArgs as Record<string, unknown>;