Image Processing MCP Server

Instructions

Implementation Reference

• tools/transform.py:138-227 (handler)

  Core handler function that performs the image resizing using PIL, with aspect ratio handling, resampling, validation, and JSON output.
  async def resize_image(image_source: str, width: int, height: int, keep_aspect_ratio: bool = True, resample: str = "LANCZOS") -> list[TextContent]: """ 调整图片大小 Args: image_source: 图片源（文件路径或base64编码数据） width: 目标宽度 height: 目标高度 keep_aspect_ratio: 是否保持宽高比 resample: 重采样方法 Returns: 调整后的图片数据 """ try: # 验证参数 if not image_source: raise ValidationError("图片源不能为空") ensure_valid_dimensions(width, height) # 验证重采样方法 resample_methods = { 'NEAREST': Image.NEAREST, 'BILINEAR': Image.BILINEAR, 'BICUBIC': Image.BICUBIC, 'LANCZOS': Image.LANCZOS } if resample.upper() not in resample_methods: raise ValidationError(f"不支持的重采样方法: {resample}") # 加载图片 image = processor.load_image(image_source) original_size = image.size # 计算目标尺寸 if keep_aspect_ratio: # 保持宽高比，计算实际尺寸 aspect_ratio = original_size[0] / original_size[1] target_aspect_ratio = width / height if aspect_ratio > target_aspect_ratio: # 以宽度为准 new_width = width new_height = int(width / aspect_ratio) else: # 以高度为准 new_height = height new_width = int(height * aspect_ratio) else: new_width, new_height = width, height # 调整大小 resized_image = image.resize( (new_width, new_height), resample_methods[resample.upper()] ) # 输出调整后的图片 output_info = processor.output_image(resized_image, f"resize_{new_width}x{new_height}") result = { "success": True, "message": f"图片大小调整成功: {original_size} -> {(new_width, new_height)}", "data": { **output_info, "original_size": original_size, "new_size": (new_width, new_height), "keep_aspect_ratio": keep_aspect_ratio, "resample_method": resample } } return [TextContent(type="text", text=json.dumps(result, ensure_ascii=False))] except ValidationError as e: error_result = { "success": False, "error": f"参数验证失败: {str(e)}" } return [TextContent(type="text", text=json.dumps(error_result, ensure_ascii=False))] except Exception as e: error_result = { "success": False, "error": f"图片大小调整失败: {str(e)}" } return [TextContent(type="text", text=json.dumps(error_result, ensure_ascii=False))]
• main.py:170-186 (registration)

  MCP tool registration using @mcp.tool() decorator, defines input schema via Annotated fields, and delegates to the core transform_resize_image implementation.
  @mcp.tool() def resize_image( image_source: Annotated[str, Field(description="图片源，可以是文件路径或base64编码的图片数据")], width: Annotated[int, Field(description="目标宽度（像素）", gt=0)], height: Annotated[int, Field(description="目标高度（像素）", gt=0)], keep_aspect_ratio: Annotated[bool, Field(description="是否保持宽高比，True时会按比例缩放", default=True)], resample: Annotated[str, Field(description="重采样算法：LANCZOS（高质量）、BILINEAR（平滑）、NEAREST（快速）", default="LANCZOS")] ) -> str: """调整图片大小""" try: result = safe_run_async(transform_resize_image(image_source, width, height, keep_aspect_ratio, resample)) return result[0].text except Exception as e: return json.dumps({ "success": False, "error": f"调整图片大小失败: {str(e)}" }, ensure_ascii=False, indent=2)
• tools/transform.py:12-47 (schema)

  Tool schema definition including input schema, properties, constraints, and defaults for resize_image.
  Tool( name="resize_image", description="调整图片大小", inputSchema={ "type": "object", "properties": { "image_data": { "type": "string", "description": "图片数据（base64编码）" }, "width": { "type": "integer", "description": "目标宽度", "minimum": 1, "maximum": 4096 }, "height": { "type": "integer", "description": "目标高度", "minimum": 1, "maximum": 4096 }, "keep_aspect_ratio": { "type": "boolean", "description": "是否保持宽高比", "default": True }, "resample": { "type": "string", "description": "重采样方法（NEAREST, BILINEAR, BICUBIC, LANCZOS）", "default": "LANCZOS" } }, "required": ["image_data", "width", "height"] } ),