chore: turn vision into capability (#679)

Fixes https://github.com/microsoft/playwright-mcp/issues/420

src/config.ts

@@ -49,7 +49,6 @@ export type CLIOptions = {
   userAgent?: string;
   userDataDir?: string;
   viewportSize?: string;
-  vision?: boolean;
 };
 
 const defaultConfig: FullConfig = {
@@ -185,7 +184,6 @@ export async function configFromCLIOptions(cliOptions: CLIOptions): Promise<Conf
       host: cliOptions.host,
     },
     capabilities: cliOptions.caps?.split(',').map((c: string) => c.trim() as ToolCapability),
-    vision: !!cliOptions.vision,
     network: {
       allowedOrigins: cliOptions.allowedOrigins,
       blockedOrigins: cliOptions.blockedOrigins,

src/connection.ts

@@ -19,7 +19,7 @@ import { CallToolRequestSchema, ListToolsRequestSchema, Tool as McpTool } from '
 import { zodToJsonSchema } from 'zod-to-json-schema';
 
 import { Context } from './context.js';
-import { snapshotTools, visionTools } from './tools.js';
+import { allTools } from './tools.js';
 import { packageJSON } from './package.js';
 
 import { FullConfig } from './config.js';
@@ -27,8 +27,7 @@ import { FullConfig } from './config.js';
 import type { BrowserContextFactory } from './browserContextFactory.js';
 
 export function createConnection(config: FullConfig, browserContextFactory: BrowserContextFactory): Connection {
-  const allTools = config.vision ? visionTools : snapshotTools;
-  const tools = allTools.filter(tool => !config.capabilities || tool.capability === 'core' || config.capabilities.includes(tool.capability));
+  const tools = allTools.filter(tool => tool.capability.startsWith('core') || config.capabilities?.includes(tool.capability));
   const context = new Context(tools, config, browserContextFactory);
   const server = new McpServer({ name: 'Playwright', version: packageJSON.version }, {
     capabilities: {

src/program.ts

@@ -14,7 +14,7 @@
  * limitations under the License.
  */
 
-import { program } from 'commander';
+import { program, Option } from 'commander';
 // @ts-ignore
 import { startTraceViewerServer } from 'playwright-core/lib/server';
 
@@ -31,7 +31,7 @@ program
     .option('--block-service-workers', 'block service workers')
     .option('--browser <browser>', 'browser or chrome channel to use, possible values: chrome, firefox, webkit, msedge.')
     .option('--browser-agent <endpoint>', 'Use browser agent (experimental).')
-    .option('--caps <caps>', 'comma-separated list of capabilities to enable, possible values: tabs, pdf, history, wait, files, install. Default is all.')
+    .option('--caps <caps>', 'comma-separated list of additional capabilities to enable, possible values: vision, pdf.')
     .option('--cdp-endpoint <endpoint>', 'CDP endpoint to connect to.')
     .option('--config <path>', 'path to the configuration file.')
     .option('--device <device>', 'device to emulate, for example: "iPhone 15"')
@@ -51,8 +51,13 @@ program
     .option('--user-agent <ua string>', 'specify user agent string')
     .option('--user-data-dir <path>', 'path to the user data directory. If not specified, a temporary directory will be created.')
     .option('--viewport-size <size>', 'specify browser viewport size in pixels, for example "1280, 720"')
-    .option('--vision', 'Run server that uses screenshots (Aria snapshots are used by default)')
+    .addOption(new Option('--vision', 'Legacy option, use --caps=vision instead').hideHelp())
     .action(async options => {
+      if (options.vision) {
+        // eslint-disable-next-line no-console
+        console.error('The --vision option is deprecated, use --caps=vision instead');
+        options.caps = 'vision';
+      }
       const config = await resolveCLIConfig(options);
       const httpServer = config.server.port !== undefined ? await startHttpServer(config.server) : undefined;
 

src/tools.ts

@@ -27,39 +27,25 @@ import pdf from './tools/pdf.js';
 import snapshot from './tools/snapshot.js';
 import tabs from './tools/tabs.js';
 import screenshot from './tools/screenshot.js';
-import vision from './tools/vision.js';
 import wait from './tools/wait.js';
+import mouse from './tools/mouse.js';
 
 import type { Tool } from './tools/tool.js';
 
-export const snapshotTools: Tool<any>[] = [
-  ...common(true),
+export const allTools: Tool<any>[] = [
+  ...common,
   ...console,
-  ...dialogs(true),
+  ...dialogs,
   ...evaluate,
-  ...files(true),
+  ...files,
   ...install,
-  ...keyboard(true),
-  ...navigate(true),
+  ...keyboard,
+  ...navigate,
   ...network,
+  ...mouse,
   ...pdf,
   ...screenshot,
   ...snapshot,
-  ...tabs(true),
-  ...wait(true),
-];
-
-export const visionTools: Tool<any>[] = [
-  ...common(false),
-  ...console,
-  ...dialogs(false),
-  ...files(false),
-  ...install,
-  ...keyboard(false),
-  ...navigate(false),
-  ...network,
-  ...pdf,
-  ...tabs(false),
-  ...vision,
-  ...wait(false),
+  ...tabs,
+  ...wait,
 ];

src/tools/common.ts

@@ -15,7 +15,7 @@
  */
 
 import { z } from 'zod';
-import { defineTool, type ToolFactory } from './tool.js';
+import { defineTool } from './tool.js';
 
 const close = defineTool({
   capability: 'core',
@@ -38,7 +38,7 @@ const close = defineTool({
   },
 });
 
-const resize: ToolFactory = captureSnapshot => defineTool({
+const resize = defineTool({
   capability: 'core',
   schema: {
     name: 'browser_resize',
@@ -66,13 +66,13 @@ const resize: ToolFactory = captureSnapshot => defineTool({
     return {
       code,
       action,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: true
     };
   },
 });
 
-export default (captureSnapshot: boolean) => [
+export default [
   close,
-  resize(captureSnapshot)
+  resize
 ];

src/tools/dialogs.ts

@@ -15,9 +15,9 @@
  */
 
 import { z } from 'zod';
-import { defineTool, type ToolFactory } from './tool.js';
+import { defineTool } from './tool.js';
 
-const handleDialog: ToolFactory = captureSnapshot => defineTool({
+const handleDialog = defineTool({
   capability: 'core',
 
   schema: {
@@ -49,7 +49,7 @@ const handleDialog: ToolFactory = captureSnapshot => defineTool({
 
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false,
     };
   },
@@ -57,6 +57,6 @@ const handleDialog: ToolFactory = captureSnapshot => defineTool({
   clearsModalState: 'dialog',
 });
 
-export default (captureSnapshot: boolean) => [
-  handleDialog(captureSnapshot),
+export default [
+  handleDialog,
 ];

src/tools/files.ts

@@ -15,10 +15,10 @@
  */
 
 import { z } from 'zod';
-import { defineTool, type ToolFactory } from './tool.js';
+import { defineTool } from './tool.js';
 
-const uploadFile: ToolFactory = captureSnapshot => defineTool({
-  capability: 'files',
+const uploadFile = defineTool({
+  capability: 'core',
 
   schema: {
     name: 'browser_file_upload',
@@ -47,13 +47,13 @@ const uploadFile: ToolFactory = captureSnapshot => defineTool({
     return {
       code,
       action,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: true,
     };
   },
   clearsModalState: 'fileChooser',
 });
 
-export default (captureSnapshot: boolean) => [
-  uploadFile(captureSnapshot),
+export default [
+  uploadFile,
 ];

src/tools/install.ts

@@ -23,7 +23,7 @@ import { defineTool } from './tool.js';
 import { fileURLToPath } from 'node:url';
 
 const install = defineTool({
-  capability: 'install',
+  capability: 'core-install',
   schema: {
     name: 'browser_install',
     title: 'Install the browser specified in the config',

src/tools/keyboard.ts

@@ -15,9 +15,13 @@
  */
 
 import { z } from 'zod';
-import { defineTool, type ToolFactory } from './tool.js';
 
-const pressKey: ToolFactory = captureSnapshot => defineTool({
+import { defineTool } from './tool.js';
+import { elementSchema } from './snapshot.js';
+import { generateLocator } from './utils.js';
+import * as javascript from '../javascript.js';
+
+const pressKey = defineTool({
   capability: 'core',
 
   schema: {
@@ -43,12 +47,61 @@ const pressKey: ToolFactory = captureSnapshot => defineTool({
     return {
       code,
       action,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: true
     };
   },
 });
 
-export default (captureSnapshot: boolean) => [
-  pressKey(captureSnapshot),
+const typeSchema = elementSchema.extend({
+  text: z.string().describe('Text to type into the element'),
+  submit: z.boolean().optional().describe('Whether to submit entered text (press Enter after)'),
+  slowly: z.boolean().optional().describe('Whether to type one character at a time. Useful for triggering key handlers in the page. By default entire text is filled in at once.'),
+});
+
+const type = defineTool({
+  capability: 'core',
+  schema: {
+    name: 'browser_type',
+    title: 'Type text',
+    description: 'Type text into editable element',
+    inputSchema: typeSchema,
+    type: 'destructive',
+  },
+
+  handle: async (context, params) => {
+    const snapshot = context.currentTabOrDie().snapshotOrDie();
+    const locator = snapshot.refLocator(params);
+
+    const code: string[] = [];
+    const steps: (() => Promise<void>)[] = [];
+
+    if (params.slowly) {
+      code.push(`// Press "${params.text}" sequentially into "${params.element}"`);
+      code.push(`await page.${await generateLocator(locator)}.pressSequentially(${javascript.quote(params.text)});`);
+      steps.push(() => locator.pressSequentially(params.text));
+    } else {
+      code.push(`// Fill "${params.text}" into "${params.element}"`);
+      code.push(`await page.${await generateLocator(locator)}.fill(${javascript.quote(params.text)});`);
+      steps.push(() => locator.fill(params.text));
+    }
+
+    if (params.submit) {
+      code.push(`// Submit text`);
+      code.push(`await page.${await generateLocator(locator)}.press('Enter');`);
+      steps.push(() => locator.press('Enter'));
+    }
+
+    return {
+      code,
+      action: () => steps.reduce((acc, step) => acc.then(step), Promise.resolve()),
+      captureSnapshot: true,
+      waitForNetwork: true,
+    };
+  },
+});
+
+export default [
+  pressKey,
+  type,
 ];

src/tools/mouse.ts

@@ -17,50 +17,14 @@
 import { z } from 'zod';
 import { defineTool } from './tool.js';
 
-import * as javascript from '../javascript.js';
-
 const elementSchema = z.object({
   element: z.string().describe('Human-readable element description used to obtain permission to interact with the element'),
 });
 
-const screenshot = defineTool({
-  capability: 'core',
+const mouseMove = defineTool({
+  capability: 'vision',
   schema: {
-    name: 'browser_screen_capture',
-    title: 'Take a screenshot',
-    description: 'Take a screenshot of the current page',
-    inputSchema: z.object({}),
-    type: 'readOnly',
-  },
-
-  handle: async context => {
-    const tab = await context.ensureTab();
-    const options = { type: 'jpeg' as 'jpeg', quality: 50, scale: 'css' as 'css' };
-
-    const code = [
-      `// Take a screenshot of the current page`,
-      `await page.screenshot(${javascript.formatObject(options)});`,
-    ];
-
-    const action = () => tab.page.screenshot(options).then(buffer => {
-      return {
-        content: [{ type: 'image' as 'image', data: buffer.toString('base64'), mimeType: 'image/jpeg' }],
-      };
-    });
-
-    return {
-      code,
-      action,
-      captureSnapshot: false,
-      waitForNetwork: false
-    };
-  },
-});
-
-const moveMouse = defineTool({
-  capability: 'core',
-  schema: {
-    name: 'browser_screen_move_mouse',
+    name: 'browser_mouse_move_xy',
     title: 'Move mouse',
     description: 'Move mouse to a given position',
     inputSchema: elementSchema.extend({
@@ -86,12 +50,12 @@ const moveMouse = defineTool({
   },
 });
 
-const click = defineTool({
-  capability: 'core',
+const mouseClick = defineTool({
+  capability: 'vision',
   schema: {
-    name: 'browser_screen_click',
+    name: 'browser_mouse_click_xy',
     title: 'Click',
-    description: 'Click left mouse button',
+    description: 'Click left mouse button at a given position',
     inputSchema: elementSchema.extend({
       x: z.number().describe('X coordinate'),
       y: z.number().describe('Y coordinate'),
@@ -121,12 +85,12 @@ const click = defineTool({
   },
 });
 
-const drag = defineTool({
-  capability: 'core',
+const mouseDrag = defineTool({
+  capability: 'vision',
   schema: {
-    name: 'browser_screen_drag',
+    name: 'browser_mouse_drag_xy',
     title: 'Drag mouse',
-    description: 'Drag left mouse button',
+    description: 'Drag left mouse button to a given position',
     inputSchema: elementSchema.extend({
       startX: z.number().describe('Start X coordinate'),
       startY: z.number().describe('Start Y coordinate'),
@@ -163,51 +127,8 @@ const drag = defineTool({
   },
 });
 
-const type = defineTool({
-  capability: 'core',
-  schema: {
-    name: 'browser_screen_type',
-    title: 'Type text',
-    description: 'Type text',
-    inputSchema: z.object({
-      text: z.string().describe('Text to type into the element'),
-      submit: z.boolean().optional().describe('Whether to submit entered text (press Enter after)'),
-    }),
-    type: 'destructive',
-  },
-
-  handle: async (context, params) => {
-    const tab = context.currentTabOrDie();
-
-    const code = [
-      `// Type ${params.text}`,
-      `await page.keyboard.type('${params.text}');`,
-    ];
-
-    const action = async () => {
-      await tab.page.keyboard.type(params.text);
-      if (params.submit)
-        await tab.page.keyboard.press('Enter');
-    };
-
-    if (params.submit) {
-      code.push(`// Submit text`);
-      code.push(`await page.keyboard.press('Enter');`);
-    }
-
-    return {
-      code,
-      action,
-      captureSnapshot: false,
-      waitForNetwork: true,
-    };
-  },
-});
-
 export default [
-  screenshot,
-  moveMouse,
-  click,
-  drag,
-  type,
+  mouseMove,
+  mouseClick,
+  mouseDrag,
 ];

src/tools/navigate.ts

@@ -15,9 +15,9 @@
  */
 
 import { z } from 'zod';
-import { defineTool, type ToolFactory } from './tool.js';
+import { defineTool } from './tool.js';
 
-const navigate: ToolFactory = captureSnapshot => defineTool({
+const navigate = defineTool({
   capability: 'core',
 
   schema: {
@@ -41,14 +41,14 @@ const navigate: ToolFactory = captureSnapshot => defineTool({
 
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false,
     };
   },
 });
 
-const goBack: ToolFactory = captureSnapshot => defineTool({
-  capability: 'history',
+const goBack = defineTool({
+  capability: 'core',
   schema: {
     name: 'browser_navigate_back',
     title: 'Go back',
@@ -67,14 +67,14 @@ const goBack: ToolFactory = captureSnapshot => defineTool({
 
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false,
     };
   },
 });
 
-const goForward: ToolFactory = captureSnapshot => defineTool({
-  capability: 'history',
+const goForward = defineTool({
+  capability: 'core',
   schema: {
     name: 'browser_navigate_forward',
     title: 'Go forward',
@@ -91,14 +91,14 @@ const goForward: ToolFactory = captureSnapshot => defineTool({
     ];
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false,
     };
   },
 });
 
-export default (captureSnapshot: boolean) => [
-  navigate(captureSnapshot),
-  goBack(captureSnapshot),
-  goForward(captureSnapshot),
+export default [
+  navigate,
+  goBack,
+  goForward,
 ];

src/tools/screenshot.ts

@@ -79,7 +79,7 @@ const screenshot = defineTool({
     return {
       code,
       action,
-      captureSnapshot: true,
+      captureSnapshot: false,
       waitForNetwork: false,
     };
   }

src/tools/snapshot.ts

@@ -41,7 +41,7 @@ const snapshot = defineTool({
   },
 });
 
-const elementSchema = z.object({
+export const elementSchema = z.object({
   element: z.string().describe('Human-readable element description used to obtain permission to interact with the element'),
   ref: z.string().describe('Exact target element reference from the page snapshot'),
 });
@@ -144,54 +144,6 @@ const hover = defineTool({
   },
 });
 
-const typeSchema = elementSchema.extend({
-  text: z.string().describe('Text to type into the element'),
-  submit: z.boolean().optional().describe('Whether to submit entered text (press Enter after)'),
-  slowly: z.boolean().optional().describe('Whether to type one character at a time. Useful for triggering key handlers in the page. By default entire text is filled in at once.'),
-});
-
-const type = defineTool({
-  capability: 'core',
-  schema: {
-    name: 'browser_type',
-    title: 'Type text',
-    description: 'Type text into editable element',
-    inputSchema: typeSchema,
-    type: 'destructive',
-  },
-
-  handle: async (context, params) => {
-    const snapshot = context.currentTabOrDie().snapshotOrDie();
-    const locator = snapshot.refLocator(params);
-
-    const code: string[] = [];
-    const steps: (() => Promise<void>)[] = [];
-
-    if (params.slowly) {
-      code.push(`// Press "${params.text}" sequentially into "${params.element}"`);
-      code.push(`await page.${await generateLocator(locator)}.pressSequentially(${javascript.quote(params.text)});`);
-      steps.push(() => locator.pressSequentially(params.text));
-    } else {
-      code.push(`// Fill "${params.text}" into "${params.element}"`);
-      code.push(`await page.${await generateLocator(locator)}.fill(${javascript.quote(params.text)});`);
-      steps.push(() => locator.fill(params.text));
-    }
-
-    if (params.submit) {
-      code.push(`// Submit text`);
-      code.push(`await page.${await generateLocator(locator)}.press('Enter');`);
-      steps.push(() => locator.press('Enter'));
-    }
-
-    return {
-      code,
-      action: () => steps.reduce((acc, step) => acc.then(step), Promise.resolve()),
-      captureSnapshot: true,
-      waitForNetwork: true,
-    };
-  },
-});
-
 const selectOptionSchema = elementSchema.extend({
   values: z.array(z.string()).describe('Array of values to select in the dropdown. This can be a single value or multiple values.'),
 });
@@ -229,6 +181,5 @@ export default [
   click,
   drag,
   hover,
-  type,
   selectOption,
 ];

src/tools/tabs.ts

@@ -15,10 +15,10 @@
  */
 
 import { z } from 'zod';
-import { defineTool, type ToolFactory } from './tool.js';
+import { defineTool } from './tool.js';
 
 const listTabs = defineTool({
-  capability: 'tabs',
+  capability: 'core-tabs',
 
   schema: {
     name: 'browser_tab_list',
@@ -44,8 +44,8 @@ const listTabs = defineTool({
   },
 });
 
-const selectTab: ToolFactory = captureSnapshot => defineTool({
-  capability: 'tabs',
+const selectTab = defineTool({
+  capability: 'core-tabs',
 
   schema: {
     name: 'browser_tab_select',
@@ -65,14 +65,14 @@ const selectTab: ToolFactory = captureSnapshot => defineTool({
 
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false
     };
   },
 });
 
-const newTab: ToolFactory = captureSnapshot => defineTool({
-  capability: 'tabs',
+const newTab = defineTool({
+  capability: 'core-tabs',
 
   schema: {
     name: 'browser_tab_new',
@@ -94,14 +94,14 @@ const newTab: ToolFactory = captureSnapshot => defineTool({
     ];
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false
     };
   },
 });
 
-const closeTab: ToolFactory = captureSnapshot => defineTool({
-  capability: 'tabs',
+const closeTab = defineTool({
+  capability: 'core-tabs',
 
   schema: {
     name: 'browser_tab_close',
@@ -120,15 +120,15 @@ const closeTab: ToolFactory = captureSnapshot => defineTool({
     ];
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false
     };
   },
 });
 
-export default (captureSnapshot: boolean) => [
+export default [
   listTabs,
-  newTab(captureSnapshot),
-  selectTab(captureSnapshot),
-  closeTab(captureSnapshot),
+  newTab,
+  selectTab,
+  closeTab,
 ];

src/tools/tool.ts

@@ -61,8 +61,6 @@ export type Tool<Input extends InputType = InputType> = {
   handle: (context: Context, params: z.output<Input>) => Promise<ToolResult>;
 };
 
-export type ToolFactory = (snapshot: boolean) => Tool<any>;
-
 export function defineTool<Input extends InputType>(tool: Tool<Input>): Tool<Input> {
   return tool;
 }

src/tools/wait.ts

@@ -15,10 +15,10 @@
  */
 
 import { z } from 'zod';
-import { defineTool, type ToolFactory } from './tool.js';
+import { defineTool } from './tool.js';
 
-const wait: ToolFactory = captureSnapshot => defineTool({
-  capability: 'wait',
+const wait = defineTool({
+  capability: 'core',
 
   schema: {
     name: 'browser_wait_for',
@@ -40,7 +40,7 @@ const wait: ToolFactory = captureSnapshot => defineTool({
 
     if (params.time) {
       code.push(`await new Promise(f => setTimeout(f, ${params.time!} * 1000));`);
-      await new Promise(f => setTimeout(f, Math.min(10000, params.time! * 1000)));
+      await new Promise(f => setTimeout(f, Math.min(30000, params.time! * 1000)));
     }
 
     const tab = context.currentTabOrDie();
@@ -59,12 +59,12 @@ const wait: ToolFactory = captureSnapshot => defineTool({
 
     return {
       code,
-      captureSnapshot,
+      captureSnapshot: true,
       waitForNetwork: false,
     };
   },
 });
 
-export default (captureSnapshot: boolean) => [
-  wait(captureSnapshot),
+export default [
+  wait,
 ];