with one click
game-audio-prompter
// An AI prompt engineering skill that reads game design documents and audio asset lists to automatically construct specialized BGM and SFX prompts for external Audio LLMs.
// An AI prompt engineering skill that reads game design documents and audio asset lists to automatically construct specialized BGM and SFX prompts for external Audio LLMs.
[HINT] Download the complete skill directory including SKILL.md and all related files
| name | game-audio-prompter |
| description | An AI prompt engineering skill that reads game design documents and audio asset lists to automatically construct specialized BGM and SFX prompts for external Audio LLMs. |
This skill acts as a specialized Game Audio Designer Agent. It translates technical mechanics and visual art styles described in Game Design Documents (GDD) into precise, well-formatted Audio Generation Prompts suitable for state-of-the-art models like Suno, Udio, ElevenLabs, or ChatGPT Advanced Voice.
When the user invokes the game-audio-prompter skill on a specific project, YOU MUST rigorously execute the following phases:
view_file to read the *_GDD.md document. Extract:
view_file to read the *_Audio_Assets.json file. Extract the full list of required audio assets, noting their IDs, Layer type (BGM vs SFX), and Loop flags.You cannot just copy the description over. You MUST translate visual/mechanic tags into Audio Envelopes and Textures using your LLM inference.
loop_flag: true MUST forcefully include the prompt constraint: "MUST be a seamless looping track, no fade-in or fade-out."Create a Markdown file named <ProjectName>_Audio_Prompt_Book.md in the project's root directory.
The generated file MUST strictly follow the structural template below.
# 🎵 [Project Name]: Game Audio Generation Prompt Book
*Tài liệu này được định dạng chuyên biệt để bạn sao chép (copy-paste) vào các Audio LLM (ElevenLabs, Udio, Suno, v.v.).*
## 1. System Definition
- **System Prompt:** "Act as a professional Game Audio Designer. You are creating the audio soundscape for a [Genre] game with a [Visual Style] aesthetic. The audio must feel [Translated Vibe]."
- **Global Constraints:** [Note any limits, like "Must be extremely short files under 50kb for Playable Ads", or "No orchestral overlaps"].
## 2. Background Music (BGM) Request
### [BGM Asset ID]
- **Gameplay Context:** [Where/When it plays]
- **Duration Constraint:** [Expected track/loop length, e.g., ~15s to 30s loop]
- **Prompt Formulation:** "[Estimated BPM], [Instruments], [Mood]. [Loop constraints]."
## 3. Sound Effects (SFX) Breakdown
*(Repeat for EVERY SFX found in the JSON - DO NOT truncate unless explicitly instructed)*
### [SFX Asset ID]
- **Gameplay Context:** [What action triggers this sound]
- **Trigger Frequency:** [Low / High / Continuous]
- **Duration Constraint:** [Strict maximum length, e.g., Max 0.5s or Max 2s]
- **Audio Texture Prompt:** "[Attack/Decay envelope], [Reverb level], [Timbre/Pitch]. [Fatigue constraints if high-frequency]."
Once the file is generated, review it structurally. Provide the user with a clickable link to the generated document and ask if they need assistance executing the generated prompts via other tools or APIs.
/finish workflow to log successful audio prompting keywords and formatting techniques into the project's learned documentation.