AI Video Generator MCP Server

Instructions

Implementation Reference

• src/index.ts:207-257 (handler)

  Handler for the 'check-video-status' tool. Determines the model, validates it, calls falClient.queue.status with the requestId, processes the response to extract status, logs, and position, and returns a JSON-formatted text content response. Handles errors appropriately.

  } else if (name === "check-video-status") {
    try {
      // Determine which model to use
      const modelName = args.model || "luma";
      const modelUrl = MODELS[modelName as keyof typeof MODELS];
      
      if (!modelUrl) {
        return {
          isError: true,
          content: [
            {
              type: "text",
              text: `Invalid model: ${modelName}. Supported models are: luma, kling`
            }
          ]
        };
      }
  
      const status: any = await falClient.queue.status(modelUrl, {
        requestId: args.request_id,
        logs: true
      });
      
      // Safely extract properties
      const statusLogs = Array.isArray(status.logs) ? status.logs : [];
      const position = status.position || status.queue_position || 0;
      
      return {
        content: [
          {
            type: "text",
            text: JSON.stringify({
              model: modelName,
              status: status.status,
              logs: statusLogs,
              position: position
            }, null, 2)
          }
        ]
      };
    } catch (error: any) {
      return {
        isError: true,
        content: [
          {
            type: "text",
            text: `Error checking video status: ${error instanceof Error ? error.message : String(error)}`
          }
        ]
      };
    }

• src/index.ts:94-113 (schema)

  Registration and schema for 'check-video-status' tool in the ListToolsRequestSchema handler. Defines name, description, and input schema requiring 'request_id' and optional 'model'.

  {
    name: "check-video-status",
    description: "Check the status of a video generation request",
    inputSchema: {
      type: "object",
      properties: {
        request_id: {
          type: "string",
          description: "The request ID to check"
        },
        model: {
          type: "string",
          enum: ["luma", "kling"],
          default: "luma",
          description: "AI model used for the request (luma=Ray2, kling=Kling)"
        }
      },
      required: ["request_id"]
    }
  }

• src/index.ts:40-116 (registration)

  The ListToolsRequestSchema handler where the 'check-video-status' tool is registered by including it in the tools array.

  server.setRequestHandler(ListToolsRequestSchema, async () => {
    return {
      tools: [
        {
          name: "generate-video",
          description: "Generate a video from text prompt and/or images using AI models (Luma or Kling)",
          inputSchema: {
            type: "object",
            properties: {
              prompt: {
                type: "string",
                description: "Text description of the desired video content"
              },
              image_url: {
                type: "string",
                description: "Initial image to start the video from (URL or base64 data URI)"
              },
              end_image_url: {
                type: "string",
                description: "Final image to end the video with (URL or base64 data URI)"
              },
              aspect_ratio: {
                type: "string",
                enum: ["16:9", "9:16", "4:3", "3:4", "21:9", "9:21"],
                default: "16:9",
                description: "Aspect ratio of the video"
              },
              resolution: {
                type: "string",
                enum: ["540p", "720p", "1080p"],
                default: "540p",
                description: "Resolution of the video (higher resolutions use more credits)"
              },
              duration: {
                type: "string",
                enum: ["5s", "9s"],
                default: "5s",
                description: "Duration of the video (9s costs 2x more)"
              },
              loop: {
                type: "boolean",
                default: false,
                description: "Whether the video should loop (blend end with beginning)"
              },
              model: {
                type: "string",
                enum: ["luma", "kling"],
                default: "luma",
                description: "AI model to use (luma=Ray2, kling=Kling)"
              }
            },
            required: ["prompt"]
          }
        },
        {
          name: "check-video-status",
          description: "Check the status of a video generation request",
          inputSchema: {
            type: "object",
            properties: {
              request_id: {
                type: "string",
                description: "The request ID to check"
              },
              model: {
                type: "string",
                enum: ["luma", "kling"],
                default: "luma",
                description: "AI model used for the request (luma=Ray2, kling=Kling)"
              }
            },
            required: ["request_id"]
          }
        }
      ]
    };
  });