surfagent 1.0.5 → 1.0.7

package/dist/api/act.d.ts

@@ -23,6 +23,7 @@ export interface ClickRequest {
     tab: string;
     selector?: string;
     text?: string;
+    waitAfter?: number;
 }
 export declare function clickElement(request: ClickRequest, options: {
     port?: number;
@@ -70,6 +71,10 @@ export declare function readPage(tabPattern: string, options: {
     host?: string;
     selector?: string;
 }): Promise<any>;
+export declare function dismissOverlays(tabPattern: string, options: {
+    port?: number;
+    host?: string;
+}): Promise<any>;
 export interface CaptchaRequest {
     tab: string;
     action: 'detect' | 'read' | 'next' | 'prev' | 'submit' | 'audio' | 'restart';

package/dist/api/act.js

@@ -135,9 +135,9 @@ export async function clickElement(request, options) {
           }
           if (!el && text) {
             const lower = text.toLowerCase();
-            const all = document.querySelectorAll('a, button, input[type="submit"], [role="button"], [onclick]');
+            const all = document.querySelectorAll('a, button, input[type="submit"], [role="button"], [role="option"], [role="menuitem"], [role="listitem"], [role="tab"], [role="link"], li[aria-label], [onclick]');
             for (const candidate of all) {
-              const t = (candidate.innerText || candidate.textContent || candidate.value || '').trim();
+              const t = (candidate.innerText || candidate.textContent || candidate.value || candidate.getAttribute('aria-label') || '').trim();
               if (t.toLowerCase().includes(lower)) { el = candidate; break; }
             }
           }
@@ -159,6 +159,10 @@ export async function clickElement(request, options) {
       `,
             returnByValue: true
         });
+        // Wait after click if requested (for page to settle after navigation/SPA route change)
+        if (request.waitAfter && request.waitAfter > 0) {
+            await new Promise(resolve => setTimeout(resolve, Math.min(request.waitAfter, 10000)));
+        }
         await client.close();
         return result.result.value;
     }
@@ -357,6 +361,67 @@ export async function readPage(tabPattern, options) {
         throw error;
     }
 }
+const DISMISS_OVERLAYS_SCRIPT = `
+(function() {
+  const dismissed = [];
+
+  // Common cookie consent button patterns (multi-language)
+  const consentPatterns = [
+    'reject all', 'reject', 'decline', 'deny',
+    'accept all', 'accept', 'godta alle', 'godta',
+    'alle ablehnen', 'ablehnen', 'tout refuser', 'refuser',
+    'rechazar todo', 'rechazar', 'rifiuta tutto', 'rifiuta',
+    'bare nødvendige', 'only necessary', 'nur notwendige',
+    'manage preferences', 'cookie settings',
+  ];
+
+  // Try cookie consent buttons
+  for (const btn of document.querySelectorAll('button, a[role="button"]')) {
+    const text = (btn.innerText || btn.textContent || '').trim().toLowerCase();
+    if (text.length > 50 || text.length < 2) continue;
+    for (const pattern of consentPatterns) {
+      if (text === pattern || text.startsWith(pattern)) {
+        btn.click();
+        dismissed.push({ type: 'cookie', text: text.substring(0, 40) });
+        break;
+      }
+    }
+    if (dismissed.length) break;
+  }
+
+  // Try closing modal dialogs (X button, close button, dismiss)
+  if (!dismissed.length) {
+    for (const btn of document.querySelectorAll('[aria-label*="Close" i], [aria-label*="Dismiss" i], [aria-label*="Lukk" i], [aria-label*="Schließen" i], [aria-label*="Fermer" i]')) {
+      const dialog = btn.closest('[role="dialog"], [role="alertdialog"], .modal, [data-overlay]');
+      if (dialog) {
+        btn.click();
+        dismissed.push({ type: 'dialog', text: btn.getAttribute('aria-label') || 'close' });
+        break;
+      }
+    }
+  }
+
+  return { dismissed, count: dismissed.length };
+})()
+`;
+export async function dismissOverlays(tabPattern, options) {
+    const port = options.port || 9222;
+    const host = options.host || 'localhost';
+    const tab = await resolveTab(tabPattern, port, host);
+    const client = await connectToTab(tab.id, port, host);
+    try {
+        const r = await client.Runtime.evaluate({
+            expression: DISMISS_OVERLAYS_SCRIPT,
+            returnByValue: true
+        });
+        await client.close();
+        return r.result.value;
+    }
+    catch (error) {
+        await client.close();
+        throw error;
+    }
+}
 const CAPTCHA_DETECT_SCRIPT = `
 (function() {
   // Find captcha iframes on the page

package/dist/api/server.js

@@ -1,7 +1,7 @@
 #!/usr/bin/env node
 import http from 'node:http';
 import { reconUrl, reconTab } from './recon.js';
-import { fillFields, clickElement, scrollPage, navigatePage, evalInTab, focusTab, readPage, captchaInteract } from './act.js';
+import { fillFields, clickElement, scrollPage, navigatePage, evalInTab, focusTab, readPage, captchaInteract, dismissOverlays } from './act.js';
 import { getAllTabs } from '../chrome/tabs.js';
 const PORT = parseInt(process.env.API_PORT || '3456', 10);
 const CDP_PORT = parseInt(process.env.CDP_PORT || '9222', 10);
@@ -16,6 +16,11 @@ function json(res, status, data) {
     res.writeHead(status, { 'Content-Type': 'application/json', 'Access-Control-Allow-Origin': '*' });
     res.end(JSON.stringify(data));
 }
+function parseBody(raw) {
+    if (!raw || !raw.trim())
+        throw new SyntaxError('Empty request body');
+    return JSON.parse(raw);
+}
 function cors(res) {
     res.writeHead(204, {
         'Access-Control-Allow-Origin': '*',
@@ -32,7 +37,7 @@ const server = http.createServer(async (req, res) => {
     try {
         // POST /recon — full page reconnaissance
         if (path === '/recon' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.url && !body.tab) {
                 return json(res, 400, { error: 'Provide "url" (to open new page) or "tab" (to recon existing tab)' });
             }
@@ -56,17 +61,20 @@ const server = http.createServer(async (req, res) => {
         }
         // POST /fill — fill form fields via CDP keystrokes
         if (path === '/fill' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.tab || !body.fields) {
                 return json(res, 400, { error: 'Provide "tab" and "fields" [{ selector, value }]' });
             }
+            if (!Array.isArray(body.fields)) {
+                return json(res, 400, { error: '"fields" must be an array of { selector, value }' });
+            }
             const start = Date.now();
             const result = await fillFields(body, { port: CDP_PORT, host: CDP_HOST });
             return json(res, 200, { ...result, _fillMs: Date.now() - start });
         }
         // POST /click — click an element
         if (path === '/click' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.tab || (!body.selector && !body.text)) {
                 return json(res, 400, { error: 'Provide "tab" and "selector" or "text"' });
             }
@@ -75,16 +83,25 @@ const server = http.createServer(async (req, res) => {
         }
         // POST /scroll — scroll a page
         if (path === '/scroll' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.tab) {
                 return json(res, 400, { error: 'Provide "tab", optional "direction" (down/up), "amount" (pixels)' });
             }
             const result = await scrollPage(body, { port: CDP_PORT, host: CDP_HOST });
             return json(res, 200, result);
         }
+        // POST /dismiss — dismiss cookie banners, modals, overlays
+        if (path === '/dismiss' && req.method === 'POST') {
+            const body = parseBody(await readBody(req));
+            if (!body.tab) {
+                return json(res, 400, { error: 'Provide "tab"' });
+            }
+            const result = await dismissOverlays(body.tab, { port: CDP_PORT, host: CDP_HOST });
+            return json(res, 200, result);
+        }
         // POST /captcha — detect and interact with captchas
         if (path === '/captcha' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.action) {
                 return json(res, 400, { error: 'Provide "action": detect, read, next, prev, submit, audio, restart' });
             }
@@ -96,7 +113,7 @@ const server = http.createServer(async (req, res) => {
         }
         // POST /read — get structured readable content from a page
         if (path === '/read' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.tab) {
                 return json(res, 400, { error: 'Provide "tab", optional "selector"' });
             }
@@ -105,7 +122,7 @@ const server = http.createServer(async (req, res) => {
         }
         // POST /focus — bring a tab to front
         if (path === '/focus' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.tab) {
                 return json(res, 400, { error: 'Provide "tab"' });
             }
@@ -114,7 +131,7 @@ const server = http.createServer(async (req, res) => {
         }
         // POST /eval — run JavaScript in a tab or iframe
         if (path === '/eval' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.tab || !body.expression) {
                 return json(res, 400, { error: 'Provide "tab" and "expression"' });
             }
@@ -123,7 +140,7 @@ const server = http.createServer(async (req, res) => {
         }
         // POST /navigate — go to url, back, or forward in same tab
         if (path === '/navigate' && req.method === 'POST') {
-            const body = JSON.parse(await readBody(req));
+            const body = parseBody(await readBody(req));
             if (!body.tab) {
                 return json(res, 400, { error: 'Provide "tab" and one of: "url", "back":true, "forward":true' });
             }
@@ -150,6 +167,15 @@ const server = http.createServer(async (req, res) => {
     catch (error) {
         const message = error instanceof Error ? error.message : String(error);
         console.error(`[${new Date().toISOString()}] Error:`, message);
+        if (error instanceof SyntaxError) {
+            return json(res, 400, { error: 'Invalid JSON: ' + message });
+        }
+        if (message.includes('Tab not found')) {
+            return json(res, 404, { error: message });
+        }
+        if (message.includes('Cannot connect to Chrome') || message.includes('ECONNREFUSED')) {
+            return json(res, 503, { error: 'Chrome not running. Start with: surfagent start' });
+        }
         json(res, 500, { error: message });
     }
 });

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "surfagent",
-  "version": "1.0.5",
+  "version": "1.0.7",
   "description": "Browser automation API for AI agents — structured page recon, form filling, clicking, and navigation via Chrome CDP",
   "keywords": [
     "ai-agent",

package/src/api/act.ts

@@ -148,6 +148,7 @@ export interface ClickRequest {
   tab: string;
   selector?: string;
   text?: string;
+  waitAfter?: number; // ms to wait after click for page to settle
 }
 
 export async function clickElement(
@@ -173,9 +174,9 @@ export async function clickElement(
           }
           if (!el && text) {
             const lower = text.toLowerCase();
-            const all = document.querySelectorAll('a, button, input[type="submit"], [role="button"], [onclick]');
+            const all = document.querySelectorAll('a, button, input[type="submit"], [role="button"], [role="option"], [role="menuitem"], [role="listitem"], [role="tab"], [role="link"], li[aria-label], [onclick]');
             for (const candidate of all) {
-              const t = (candidate.innerText || candidate.textContent || candidate.value || '').trim();
+              const t = (candidate.innerText || candidate.textContent || candidate.value || candidate.getAttribute('aria-label') || '').trim();
               if (t.toLowerCase().includes(lower)) { el = candidate; break; }
             }
           }
@@ -198,6 +199,11 @@ export async function clickElement(
       returnByValue: true
     });
 
+    // Wait after click if requested (for page to settle after navigation/SPA route change)
+    if (request.waitAfter && request.waitAfter > 0) {
+      await new Promise<void>(resolve => setTimeout(resolve, Math.min(request.waitAfter!, 10000)));
+    }
+
     await client.close();
     return result.result.value as any;
   } catch (error) {
@@ -434,6 +440,73 @@ export async function readPage(
   }
 }
 
+const DISMISS_OVERLAYS_SCRIPT = `
+(function() {
+  const dismissed = [];
+
+  // Common cookie consent button patterns (multi-language)
+  const consentPatterns = [
+    'reject all', 'reject', 'decline', 'deny',
+    'accept all', 'accept', 'godta alle', 'godta',
+    'alle ablehnen', 'ablehnen', 'tout refuser', 'refuser',
+    'rechazar todo', 'rechazar', 'rifiuta tutto', 'rifiuta',
+    'bare nødvendige', 'only necessary', 'nur notwendige',
+    'manage preferences', 'cookie settings',
+  ];
+
+  // Try cookie consent buttons
+  for (const btn of document.querySelectorAll('button, a[role="button"]')) {
+    const text = (btn.innerText || btn.textContent || '').trim().toLowerCase();
+    if (text.length > 50 || text.length < 2) continue;
+    for (const pattern of consentPatterns) {
+      if (text === pattern || text.startsWith(pattern)) {
+        btn.click();
+        dismissed.push({ type: 'cookie', text: text.substring(0, 40) });
+        break;
+      }
+    }
+    if (dismissed.length) break;
+  }
+
+  // Try closing modal dialogs (X button, close button, dismiss)
+  if (!dismissed.length) {
+    for (const btn of document.querySelectorAll('[aria-label*="Close" i], [aria-label*="Dismiss" i], [aria-label*="Lukk" i], [aria-label*="Schließen" i], [aria-label*="Fermer" i]')) {
+      const dialog = btn.closest('[role="dialog"], [role="alertdialog"], .modal, [data-overlay]');
+      if (dialog) {
+        btn.click();
+        dismissed.push({ type: 'dialog', text: btn.getAttribute('aria-label') || 'close' });
+        break;
+      }
+    }
+  }
+
+  return { dismissed, count: dismissed.length };
+})()
+`;
+
+export async function dismissOverlays(
+  tabPattern: string,
+  options: { port?: number; host?: string }
+): Promise<any> {
+  const port = options.port || 9222;
+  const host = options.host || 'localhost';
+
+  const tab = await resolveTab(tabPattern, port, host);
+  const client = await connectToTab(tab.id, port, host);
+
+  try {
+    const r = await client.Runtime.evaluate({
+      expression: DISMISS_OVERLAYS_SCRIPT,
+      returnByValue: true
+    });
+    await client.close();
+    return r.result.value;
+  } catch (error) {
+    await client.close();
+    throw error;
+  }
+}
+
 const CAPTCHA_DETECT_SCRIPT = `
 (function() {
   // Find captcha iframes on the page

package/src/api/server.ts

@@ -2,7 +2,7 @@
 
 import http from 'node:http';
 import { reconUrl, reconTab } from './recon.js';
-import { fillFields, clickElement, scrollPage, navigatePage, evalInTab, focusTab, readPage, captchaInteract } from './act.js';
+import { fillFields, clickElement, scrollPage, navigatePage, evalInTab, focusTab, readPage, captchaInteract, dismissOverlays } from './act.js';
 import { getAllTabs } from '../chrome/tabs.js';
 
 const PORT = parseInt(process.env.API_PORT || '3456', 10);
@@ -27,6 +27,11 @@ function json(res: http.ServerResponse, status: number, data: any) {
   res.end(JSON.stringify(data));
 }
 
+function parseBody(raw: string): any {
+  if (!raw || !raw.trim()) throw new SyntaxError('Empty request body');
+  return JSON.parse(raw);
+}
+
 function cors(res: http.ServerResponse) {
   res.writeHead(204, {
     'Access-Control-Allow-Origin': '*',
@@ -45,7 +50,7 @@ const server = http.createServer(async (req, res) => {
   try {
     // POST /recon — full page reconnaissance
     if (path === '/recon' && req.method === 'POST') {
-      const body: RequestBody = JSON.parse(await readBody(req));
+      const body: RequestBody = parseBody(await readBody(req));
 
       if (!body.url && !body.tab) {
         return json(res, 400, { error: 'Provide "url" (to open new page) or "tab" (to recon existing tab)' });
@@ -73,10 +78,13 @@ const server = http.createServer(async (req, res) => {
 
     // POST /fill — fill form fields via CDP keystrokes
     if (path === '/fill' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.tab || !body.fields) {
         return json(res, 400, { error: 'Provide "tab" and "fields" [{ selector, value }]' });
       }
+      if (!Array.isArray(body.fields)) {
+        return json(res, 400, { error: '"fields" must be an array of { selector, value }' });
+      }
       const start = Date.now();
       const result = await fillFields(body, { port: CDP_PORT, host: CDP_HOST });
       return json(res, 200, { ...result, _fillMs: Date.now() - start });
@@ -84,7 +92,7 @@ const server = http.createServer(async (req, res) => {
 
     // POST /click — click an element
     if (path === '/click' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.tab || (!body.selector && !body.text)) {
         return json(res, 400, { error: 'Provide "tab" and "selector" or "text"' });
       }
@@ -94,7 +102,7 @@ const server = http.createServer(async (req, res) => {
 
     // POST /scroll — scroll a page
     if (path === '/scroll' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.tab) {
         return json(res, 400, { error: 'Provide "tab", optional "direction" (down/up), "amount" (pixels)' });
       }
@@ -102,9 +110,19 @@ const server = http.createServer(async (req, res) => {
       return json(res, 200, result);
     }
 
+    // POST /dismiss — dismiss cookie banners, modals, overlays
+    if (path === '/dismiss' && req.method === 'POST') {
+      const body = parseBody(await readBody(req));
+      if (!body.tab) {
+        return json(res, 400, { error: 'Provide "tab"' });
+      }
+      const result = await dismissOverlays(body.tab, { port: CDP_PORT, host: CDP_HOST });
+      return json(res, 200, result);
+    }
+
     // POST /captcha — detect and interact with captchas
     if (path === '/captcha' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.action) {
         return json(res, 400, { error: 'Provide "action": detect, read, next, prev, submit, audio, restart' });
       }
@@ -117,7 +135,7 @@ const server = http.createServer(async (req, res) => {
 
     // POST /read — get structured readable content from a page
     if (path === '/read' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.tab) {
         return json(res, 400, { error: 'Provide "tab", optional "selector"' });
       }
@@ -127,7 +145,7 @@ const server = http.createServer(async (req, res) => {
 
     // POST /focus — bring a tab to front
     if (path === '/focus' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.tab) {
         return json(res, 400, { error: 'Provide "tab"' });
       }
@@ -137,7 +155,7 @@ const server = http.createServer(async (req, res) => {
 
     // POST /eval — run JavaScript in a tab or iframe
     if (path === '/eval' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.tab || !body.expression) {
         return json(res, 400, { error: 'Provide "tab" and "expression"' });
       }
@@ -147,7 +165,7 @@ const server = http.createServer(async (req, res) => {
 
     // POST /navigate — go to url, back, or forward in same tab
     if (path === '/navigate' && req.method === 'POST') {
-      const body = JSON.parse(await readBody(req));
+      const body = parseBody(await readBody(req));
       if (!body.tab) {
         return json(res, 400, { error: 'Provide "tab" and one of: "url", "back":true, "forward":true' });
       }
@@ -175,6 +193,16 @@ const server = http.createServer(async (req, res) => {
   } catch (error) {
     const message = error instanceof Error ? error.message : String(error);
     console.error(`[${new Date().toISOString()}] Error:`, message);
+
+    if (error instanceof SyntaxError) {
+      return json(res, 400, { error: 'Invalid JSON: ' + message });
+    }
+    if (message.includes('Tab not found')) {
+      return json(res, 404, { error: message });
+    }
+    if (message.includes('Cannot connect to Chrome') || message.includes('ECONNREFUSED')) {
+      return json(res, 503, { error: 'Chrome not running. Start with: surfagent start' });
+    }
     json(res, 500, { error: message });
   }
 });