@nbakka/mcp-appium 4.0.1 → 4.0.2

package/lib/server.js

@@ -177,7 +177,7 @@ tool(
      return `Screen elements (${elements.length} found): ${JSON.stringify(elements, null, 2)}`;
  });
 
-    toool("mobile_press_button", "Press a button on device", {
+    tool("mobile_press_button", "Press a button on device", {
         button: zod_1.z.string().describe("The button to press. Supported buttons: BACK (android only), HOME, VOLUME_UP, VOLUME_DOWN, ENTER, DPAD_CENTER (android tv only), DPAD_UP (android tv only), DPAD_DOWN (android tv only), DPAD_LEFT (android tv only), DPAD_RIGHT (android tv only)"),
     }, async ({ button }) => {
         requireRobot();
@@ -230,33 +230,49 @@ tool(
 
     tool(
       "mobile_tap_by_text",
-      "Tap an element on screen by its displayed text or accessibility label using ADB tap",
+      "Tap an element on screen by any attribute: text, accessibilityId (content-desc), or id (resource-id). Searches all attributes automatically.",
       {
-        text: zod_1.z.string().describe("The exact text or label of the element to tap"),
+        value: zod_1.z.string().describe("The value to search for - can be text, accessibilityId, or id"),
       },
-      async ({ text }) => {
-        if (!text) throw new Error("Input text is required");
+      async ({ value }) => {
+        if (!value) throw new Error("Input value is required");
 
         requireRobot(); // ensure robot instance available
-        const elements = await robot.getElementsOnScreen();
+        const elements = await robot.getSimplifiedElements();
 
-        // Find element by exact match on text or label
+        // Find element by matching any attribute (text, accessibilityId, or id)
         const element = elements.find(
-          el => el.text === text || el.label === text
+          el => el.text === value || el.accessibilityId === value || el.id === value
         );
 
-        if (!element) throw new Error(`Element with text or label "${text}" not found`);
-
+        if (!element) throw new Error(`Element with value "${value}" not found in text, accessibilityId, or id`);
+
+        // Parse bounds to get coordinates
+        // Bounds format: "[x1,y1][x2,y2]"
+        if (!element.bounds) throw new Error("Element has no bounds information");
+        
+        const boundsMatch = element.bounds.match(/\[(\d+),(\d+)\]\[(\d+),(\d+)\]/);
+        if (!boundsMatch) throw new Error("Invalid bounds format");
+        
+        const x1 = parseInt(boundsMatch[1]);
+        const y1 = parseInt(boundsMatch[2]);
+        const x2 = parseInt(boundsMatch[3]);
+        const y2 = parseInt(boundsMatch[4]);
+        
         // Calculate center coordinates
-        const rect = element.rect;
-        const x = Math.floor(rect.x + rect.width / 2);
-        const y = Math.floor(rect.y + rect.height / 2);
+        const x = Math.floor((x1 + x2) / 2);
+        const y = Math.floor((y1 + y2) / 2);
+
+        // Execute tap using robot (works for both Android and iOS)
+        await robot.tap(x, y);
 
-        // Execute adb tap
-        const { execSync } = require("child_process");
-        execSync(`adb shell input tap ${x} ${y}`);
+        // Determine which attribute matched
+        let matchedBy = '';
+        if (element.text === value) matchedBy = 'text';
+        else if (element.accessibilityId === value) matchedBy = 'accessibilityId';
+        else if (element.id === value) matchedBy = 'id';
 
-        return `Tapped element with text/label "${text}" at (${x},${y})`;
+        return `Tapped element by ${matchedBy} "${value}" at (${x},${y})`;
       }
     );
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@nbakka/mcp-appium",
-  "version": "4.0.1",
+  "version": "4.0.2",
   "description": "Appium MCP",
   "engines": {
     "node": ">=18"