What problem does it solve? Writing effective prompts for AI music generation models is difficult: vague prompts produce inconsistent results, and requests longer than the model's ~184-second limit fail or get truncated. This Skill provides a systematic framework for translating a musical vision into precise, structured prompts that the model can execute reliably. ## Core Features & Use Cases - 9-Dimension Prompt Framework: Combines genre, tempo, key, mood, instrumentation, density, arrangement, soundscape, and production quality into a single cohesive prompt string. - Duration-Aware Strategy: Enforces single-call generation for tracks up to 180 seconds and applies a multi-clip continuity strategy with timestamp cues and intensity markers for longer pieces. - In-Text Negative Prompting & Control Emulation: Excludes unwanted elements (drums, vocals, fast tempos) and emulates advanced parameters like density, brightness, and mute-drums through natural language. - Use Case: A user requests a 6-minute cinematic score. The Skill plans the full arrangement, locks the song's DNA (genre, BPM, production) across clips, aligns clip boundaries via precise state matching, and finishes with ffmpeg crossfades aligned to the beat grid. ## Quick Start Ask the AI to read this skill and then generate a 60-second instrumental lo-fi track at 80 BPM with a nostalgic mood and a timestamped arrangement breakdown.
