brave-real-browser-mcp-server 2.12.11 → 2.13.0

package/dist/handlers/interaction-handlers.js

@@ -167,6 +167,68 @@ export async function handleType(args) {
         }, 'Failed to type text');
     });
 }
+// Press key handler
+export async function handlePressKey(args) {
+    return await withWorkflowValidation('press_key', args, async () => {
+        return await withErrorHandling(async () => {
+            const pageInstance = getPageInstance();
+            if (!pageInstance) {
+                throw new Error('Browser not initialized. Call browser_init first.');
+            }
+            const { key, selector, modifiers = [] } = args;
+            try {
+                // If selector is provided, focus on element first
+                if (selector) {
+                    const elementResult = await selfHealingLocators.findElementWithFallbacks(pageInstance, selector);
+                    if (elementResult) {
+                        await elementResult.element.focus();
+                    }
+                    else {
+                        console.warn(`Element not found: ${selector}, pressing key on current focus`);
+                    }
+                }
+                // Press key with modifiers if specified
+                if (modifiers.length > 0) {
+                    // Hold down modifiers
+                    for (const modifier of modifiers) {
+                        await pageInstance.keyboard.down(modifier);
+                    }
+                    // Press the main key
+                    await pageInstance.keyboard.press(key);
+                    // Release modifiers
+                    for (const modifier of modifiers.reverse()) {
+                        await pageInstance.keyboard.up(modifier);
+                    }
+                    const modifierStr = modifiers.join('+');
+                    return {
+                        content: [
+                            {
+                                type: 'text',
+                                text: `Pressed key combination: ${modifierStr}+${key}${selector ? ` on element: ${selector}` : ''}\n\n✅ Keyboard interaction completed successfully`,
+                            },
+                        ],
+                    };
+                }
+                else {
+                    // Simple key press without modifiers
+                    await pageInstance.keyboard.press(key);
+                    return {
+                        content: [
+                            {
+                                type: 'text',
+                                text: `Pressed key: ${key}${selector ? ` on element: ${selector}` : ''}\n\n✅ Keyboard interaction completed successfully`,
+                            },
+                        ],
+                    };
+                }
+            }
+            catch (pressError) {
+                throw new Error(`Failed to press key: ${key}. ` +
+                    `Error: ${pressError instanceof Error ? pressError.message : String(pressError)}`);
+            }
+        }, 'Failed to press key');
+    });
+}
 // Solve captcha handler
 export async function handleSolveCaptcha(args) {
     return await withErrorHandling(async () => {

package/dist/mcp-server.js

@@ -5,7 +5,7 @@ import { validateMCPResponse } from './mcp-response-validator.js';
 // Import handlers
 import { handleBrowserInit, handleBrowserClose } from './handlers/browser-handlers.js';
 import { handleNavigate, handleWait } from './handlers/navigation-handlers.js';
-import { handleClick, handleType, handleSolveCaptcha, handleRandomScroll } from './handlers/interaction-handlers.js';
+import { handleClick, handleType, handlePressKey, handleSolveCaptcha, handleRandomScroll } from './handlers/interaction-handlers.js';
 import { handleGetContent, handleFindSelector } from './handlers/content-handlers.js';
 import { handleSaveContentAsMarkdown } from './handlers/file-handlers.js';
 import { handleScrapeTable, handleExtractList, handleExtractJSON, handleScrapeMetaTags, handleExtractSchema } from './handlers/data-extraction-handlers.js';
@@ -54,6 +54,9 @@ export async function createMcpServer() {
                 case TOOL_NAMES.TYPE:
                     result = await handleType(args);
                     break;
+                case TOOL_NAMES.PRESS_KEY:
+                    result = await handlePressKey(args);
+                    break;
                 case TOOL_NAMES.WAIT:
                     result = await handleWait(args);
                     break;

package/dist/tool-definitions.js

@@ -172,6 +172,33 @@ export const TOOLS = [
             required: ['selector', 'text'],
         },
     },
+    {
+        name: 'press_key',
+        description: 'Press keyboard keys (Enter, Tab, Escape, Arrow keys, etc.) - Useful for form submission, navigation, and keyboard shortcuts',
+        inputSchema: {
+            type: 'object',
+            properties: {
+                key: {
+                    type: 'string',
+                    description: 'Key to press. Common keys: Enter, Tab, Escape, Backspace, Delete, ArrowUp, ArrowDown, ArrowLeft, ArrowRight, Space, PageUp, PageDown, Home, End, F1-F12',
+                    enum: ['Enter', 'Tab', 'Escape', 'Backspace', 'Delete', 'Space', 'ArrowUp', 'ArrowDown', 'ArrowLeft', 'ArrowRight', 'PageUp', 'PageDown', 'Home', 'End', 'F1', 'F2', 'F3', 'F4', 'F5', 'F6', 'F7', 'F8', 'F9', 'F10', 'F11', 'F12'],
+                },
+                selector: {
+                    type: 'string',
+                    description: 'Optional: CSS selector of element to focus before pressing key',
+                },
+                modifiers: {
+                    type: 'array',
+                    description: 'Optional: Modifier keys to hold (Ctrl, Shift, Alt, Meta)',
+                    items: {
+                        type: 'string',
+                        enum: ['Control', 'Shift', 'Alt', 'Meta'],
+                    },
+                },
+            },
+            required: ['key'],
+        },
+    },
     {
         name: 'wait',
         description: 'Wait for various conditions',
@@ -1480,6 +1507,7 @@ export const TOOL_NAMES = {
     GET_CONTENT: 'get_content',
     CLICK: 'click',
     TYPE: 'type',
+    PRESS_KEY: 'press_key',
     WAIT: 'wait',
     BROWSER_CLOSE: 'browser_close',
     SOLVE_CAPTCHA: 'solve_captcha',
@@ -1554,6 +1582,6 @@ export const TOOL_NAMES = {
 export const TOOL_CATEGORIES = {
     BROWSER_MANAGEMENT: [TOOL_NAMES.BROWSER_INIT, TOOL_NAMES.BROWSER_CLOSE],
     NAVIGATION: [TOOL_NAMES.NAVIGATE, TOOL_NAMES.WAIT],
-    INTERACTION: [TOOL_NAMES.CLICK, TOOL_NAMES.TYPE, TOOL_NAMES.SOLVE_CAPTCHA, TOOL_NAMES.RANDOM_SCROLL],
+    INTERACTION: [TOOL_NAMES.CLICK, TOOL_NAMES.TYPE, TOOL_NAMES.PRESS_KEY, TOOL_NAMES.SOLVE_CAPTCHA, TOOL_NAMES.RANDOM_SCROLL],
     CONTENT: [TOOL_NAMES.GET_CONTENT, TOOL_NAMES.FIND_SELECTOR, TOOL_NAMES.SAVE_CONTENT_AS_MARKDOWN],
 };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "brave-real-browser-mcp-server",
-  "version": "2.12.11",
+  "version": "2.13.0",
   "description": "Universal AI IDE MCP Server - Auto-detects and supports all AI IDEs (Claude Desktop, Cursor, Windsurf, Cline, Zed, VSCode, Qoder AI, etc.) with Brave browser automation",
   "type": "module",
   "main": "dist/index.js",
@@ -33,6 +33,8 @@
     "test:integration": "vitest run test/integration",
     "test:e2e": "vitest run --config vitest.e2e.config.ts",
     "test:e2e:watch": "vitest --config vitest.e2e.config.ts",
+    "test:visual": "vitest run test/e2e/visual-browser.test.ts",
+    "test:visual:watch": "vitest test/e2e/visual-browser.test.ts",
     "test:brave:cleanup": "taskkill /F /IM brave.exe || pkill -f brave || true"
   },
   "dependencies": {