MCP Audio Tweaker

Instructions

Implementation Reference

• src/tools/index.ts:463-499 (handler)

  Main execution handler for the 'process_audio_file' tool. Parses input using Zod schema, attempts basic processor, falls back to advanced processor if validation fails, and returns JSON result.
  case 'process_audio_file': { try { const input = ProcessAudioFileInputSchema.parse(args); const result = await audioProcessor.processAudioFile( input.inputFile, input.outputFile, input.operations, (args as any).overwrite || false ); return { content: [ { type: 'text', text: JSON.stringify(result, null, 2) } ] }; } catch (validationError) { // If validation fails, try with the advanced processor const result = await advancedProcessor.processAudioFile( (args as any).inputFile, (args as any).outputFile, (args as any).operations, (args as any).overwrite || false ); return { content: [ { type: 'text', text: JSON.stringify(result, null, 2) } ] }; } }
• src/schemas/validation.ts:111-126 (schema)

  Zod validation schema for process_audio_file inputs, including file paths, operations for volume, format, and effects.
  export const ProcessAudioFileInputSchema = z.object({ inputFile: FilePathSchema, outputFile: z.string().min(1), operations: AudioOperationsSchema }).transform((data) => { // Transform string values to numbers for format operations if (data.operations.format) { if (data.operations.format.sampleRate && typeof data.operations.format.sampleRate === 'string') { data.operations.format.sampleRate = parseInt(data.operations.format.sampleRate) as any; } if (data.operations.format.channels && typeof data.operations.format.channels === 'string') { data.operations.format.channels = parseInt(data.operations.format.channels) as any; } } return data; });
• src/tools/index.ts:19-85 (registration)

  MCP Tool registration object defining name, description, and JSON input schema for 'process_audio_file'.
  export const processAudioFileTool: Tool = { name: 'process_audio_file', description: 'Apply audio processing operations to a single file using FFmpeg', inputSchema: { type: 'object', properties: { inputFile: { type: 'string', description: 'Path to input audio file' }, outputFile: { type: 'string', description: 'Path for output file' }, operations: { type: 'object', description: 'Audio processing operations to apply', properties: { volume: { type: 'object', properties: { adjust: { type: 'number', minimum: -60, maximum: 20 }, normalize: { type: 'boolean' }, targetLUFS: { type: 'number' } } }, format: { type: 'object', properties: { sampleRate: { type: 'number', enum: [8000, 16000, 22050, 44100, 48000, 96000, 192000] }, bitrate: { type: 'number', minimum: 64, maximum: 320 }, channels: { type: 'number', enum: [1, 2, 6, 8] }, codec: { type: 'string', enum: ['pcm', 'mp3', 'aac', 'vorbis', 'flac'] } } }, effects: { type: 'object', properties: { fadeIn: { type: 'number', minimum: 0 }, fadeOut: { type: 'number', minimum: 0 }, trim: { type: 'object', properties: { start: { type: 'number', minimum: 0 }, end: { type: 'number', minimum: 0 } } }, loop: { type: 'object', properties: { enabled: { type: 'boolean' }, count: { type: 'number', minimum: 1 } } } } } } }, overwrite: { type: 'boolean', description: 'Whether to overwrite existing output files', default: false } }, required: ['inputFile', 'outputFile', 'operations'] } };
• src/services/base-audio-processor.ts:34-79 (helper)

  Base implementation of processAudioFile method that validates inputs, builds FFmpeg command with operations, executes processing, and returns result. Used by both basic and advanced processors.
  async processAudioFile( inputFile: string, outputFile: string, operations: AudioOperations, overwrite: boolean = false ): Promise<ProcessingResult> { const startTime = Date.now(); try { // Validate input await validateInputFile(inputFile); await ensureOutputDirectory(outputFile); await handleExistingOutput(outputFile, overwrite); // Create FFmpeg command const command = createFFmpegCommand(inputFile); // Apply operations this.applyOperationsToCommand(command, operations); // Execute command await executeFFmpegCommand(command, outputFile); const processingTime = Date.now() - startTime; return { success: true, inputFile, outputFile, processingTime, operations }; } catch (error) { const processingTime = Date.now() - startTime; return { success: false, inputFile, outputFile, processingTime, operations, error: error instanceof Error ? error.message : 'Unknown error' }; } }
• src/services/advanced-audio-processor.ts:194-202 (helper)

  Overridden applyOperationsToCommand in advanced processor that adds support for advanced effects like pitch, tempo, spectral, dynamics, and spatial processing beyond base capabilities.
  protected applyOperationsToCommand(command: any, operations: AudioOperations): void { // Apply base operations first (volume, format, basic effects) super.applyOperationsToCommand(command, operations); // Apply advanced operations if (operations.advanced) { this.applyAdvancedOperations(command, operations.advanced); } }