@agentrhq/webcmd 0.1.2 → 0.2.1

package/dist/src/release-notes.js

@@ -0,0 +1,99 @@
+export const RELEASE_NOTE_SECTIONS = ['Highlights', 'Improvements', 'Fixes', 'Contributors', 'Reverts'];
+const SQUASH_MERGE_PR_NUMBER_PATTERN = /\(#(?<number>\d+)\)\s*$/;
+const MERGE_COMMIT_PR_NUMBER_PATTERN = /^Merge pull request #(?<number>\d+) /;
+const RELEASE_PLEASE_TITLE_PATTERN = /^chore(?:\([^)]+\))?: release(?:\s|$)/;
+function normalizeHandle(handle) {
+    const trimmed = handle.trim();
+    return trimmed.startsWith('@') ? trimmed.slice(1) : trimmed;
+}
+function uniqueSortedHandles(handles) {
+    return [...new Set(handles.map(normalizeHandle).filter(Boolean))].sort((left, right) => left.localeCompare(right));
+}
+function formatSectionContent(content) {
+    const trimmed = content?.trim();
+    return trimmed && trimmed.length > 0 ? trimmed : 'None.';
+}
+export function extractPullRequestNumber(message) {
+    const firstLine = message.split(/\r?\n/, 1)[0] ?? message;
+    const squashMatch = firstLine.match(SQUASH_MERGE_PR_NUMBER_PATTERN);
+    if (squashMatch?.groups?.number) {
+        return Number(squashMatch.groups.number);
+    }
+    const mergeMatch = firstLine.match(MERGE_COMMIT_PR_NUMBER_PATTERN);
+    if (mergeMatch?.groups?.number) {
+        return Number(mergeMatch.groups.number);
+    }
+    return null;
+}
+export function filterReleasePullRequests(prs) {
+    return prs.filter((pr) => {
+        const title = pr.title.trim().toLowerCase();
+        if (title.startsWith('release:'))
+            return false;
+        if (RELEASE_PLEASE_TITLE_PATTERN.test(title))
+            return false;
+        return !pr.labels.some((label) => {
+            const name = label.name.trim().toLowerCase();
+            return name === 'release' || name === 'skip-changelog';
+        });
+    });
+}
+function parseReleaseNoteSections(raw) {
+    const sections = {};
+    let currentSection = null;
+    for (const line of raw.split(/\r?\n/)) {
+        const headingMatch = line.match(/^##\s+(.+?)\s*$/);
+        if (headingMatch) {
+            const section = RELEASE_NOTE_SECTIONS.find((candidate) => candidate.toLowerCase() === headingMatch[1].toLowerCase()) ?? null;
+            currentSection = section;
+            if (currentSection)
+                sections[currentSection] = sections[currentSection] ?? [];
+            continue;
+        }
+        if (currentSection) {
+            sections[currentSection] ??= [];
+            sections[currentSection].push(line);
+        }
+    }
+    return sections;
+}
+export function normalizeReleaseNotes(raw, contributors) {
+    const sections = parseReleaseNoteSections(raw);
+    const normalizedContributors = uniqueSortedHandles(contributors).map((handle) => `- @${handle}`);
+    return RELEASE_NOTE_SECTIONS.map((section) => {
+        if (section === 'Contributors') {
+            const content = normalizedContributors.length > 0 ? normalizedContributors.join('\n') : 'None.';
+            return `## ${section}\n${content}`;
+        }
+        const content = formatSectionContent(sections[section]?.join('\n'));
+        return `## ${section}\n${content}`;
+    }).join('\n\n');
+}
+function formatPullRequest(pr) {
+    const author = pr.author?.login ?? 'unknown';
+    const labels = pr.labels.length > 0 ? pr.labels.map((label) => label.name).join(', ') : 'None';
+    const files = pr.files.length > 0 ? pr.files.map((file) => file.path).join('\n') : 'None';
+    const body = pr.body?.trim() ? pr.body.trim() : 'None';
+    const diff = pr.diff?.trim() ? pr.diff.trim() : 'None';
+    return [
+        `PR #${pr.number}: ${pr.title}`,
+        `Author: @${author}`,
+        `Labels: ${labels}`,
+        `Files:\n${files}`,
+        `Body:\n${body}`,
+        `Diff:\n${diff}`,
+    ].join('\n');
+}
+export function buildReleaseNotesPrompt(context) {
+    const prSummaries = context.pullRequests.map(formatPullRequest).join('\n\n');
+    return [
+        `Write user-facing release notes for ${context.tag}.`,
+        `Release range: ${context.previousTag}...${context.currentRef}.`,
+        'Use only the supplied pull requests below. Do not invent changes or pull in information from elsewhere.',
+        `Required sections: ${RELEASE_NOTE_SECTIONS.map((section) => `## ${section}`).join(', ')}.`,
+        'Each section must be present in the final notes.',
+        '',
+        `Pull requests included for this release:`,
+        prSummaries,
+    ].join('\n');
+}

package/dist/src/release-notes.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/src/release-notes.test.js

@@ -0,0 +1,61 @@
+import { describe, expect, it } from 'vitest';
+import { RELEASE_NOTE_SECTIONS, buildReleaseNotesPrompt, extractPullRequestNumber, filterReleasePullRequests, normalizeReleaseNotes, } from './release-notes.js';
+describe('release notes helpers', () => {
+    it('extracts PR numbers from squash and merge commit messages', () => {
+        expect(extractPullRequestNumber('feat: add release notes (#123)')).toBe(123);
+        expect(extractPullRequestNumber('feat: add release notes (#124)\n\nExpanded body')).toBe(124);
+        expect(extractPullRequestNumber('Merge pull request #456 from agentrhq/example')).toBe(456);
+        expect(extractPullRequestNumber('docs: plain commit without pr')).toBeNull();
+    });
+    it('filters release and skip-changelog pull requests', () => {
+        const prs = [
+            { number: 1, title: 'feat: browser polish', author: { login: 'alice' }, labels: [], files: [], url: 'https://example.com/1' },
+            { number: 2, title: 'release: 0.2.0', author: { login: 'bot' }, labels: [], files: [], url: 'https://example.com/2' },
+            { number: 3, title: 'chore: generated updates', author: { login: 'ci' }, labels: [{ name: 'skip-changelog' }], files: [], url: 'https://example.com/3' },
+            { number: 4, title: 'chore: release prep', author: { login: 'ci' }, labels: [{ name: 'release' }], files: [], url: 'https://example.com/4' },
+            { number: 5, title: 'chore(main): release webcmd 0.2.0', author: { login: 'release-please' }, labels: [], files: [], url: 'https://example.com/5' },
+        ];
+        expect(filterReleasePullRequests(prs).map((pr) => pr.number)).toEqual([1]);
+    });
+    it('normalizes all required sections and fills empty sections with None', () => {
+        const raw = [
+            '## Highlights',
+            '- Better release notes.',
+            '',
+            '## Fixes',
+            '- Fixed release fallback.',
+        ].join('\n');
+        const normalized = normalizeReleaseNotes(raw, ['alice', 'bob']);
+        for (const section of RELEASE_NOTE_SECTIONS) {
+            expect(normalized).toContain(`## ${section}`);
+        }
+        expect(normalized).toContain('## Improvements\nNone.');
+        expect(normalized).toContain('## Contributors\n- @alice\n- @bob');
+        expect(normalized).toContain('## Reverts\nNone.');
+    });
+    it('builds a prompt grounded in the exact release range and PR list', () => {
+        const context = {
+            tag: 'v0.2.0',
+            previousTag: 'v0.1.1',
+            currentRef: 'abc123',
+            pullRequests: [
+                {
+                    number: 42,
+                    title: 'feat: add docs scaffold',
+                    body: 'Adds docs structure.',
+                    author: { login: 'alice' },
+                    labels: [{ name: 'feature' }],
+                    files: [{ path: 'docs/docs.json' }],
+                    diff: 'diff --git a/docs/docs.json b/docs/docs.json',
+                    url: 'https://github.com/agentrhq/webcmd/pull/42',
+                },
+            ],
+        };
+        const prompt = buildReleaseNotesPrompt(context);
+        expect(prompt).toContain('v0.1.1...abc123');
+        expect(prompt).toContain('PR #42: feat: add docs scaffold');
+        expect(prompt).toContain('docs/docs.json');
+        expect(prompt).toContain('## Highlights');
+        expect(prompt).toContain('## Reverts');
+    });
+});

package/dist/src/runtime-identity.test.js

@@ -1,6 +1,5 @@
 import { describe, expect, it } from 'vitest';
 import * as path from 'node:path';
-import { unsupportedDaemonPortEnvMessage } from './constants.js';
 import { getUserWebcmdDir, getUserClisDir, getPluginsDir } from './discovery.js';
 describe('webcmd runtime identity', () => {
     it('uses webcmd runtime directories', async () => {
@@ -8,11 +7,4 @@ describe('webcmd runtime identity', () => {
         expect(getUserClisDir('/home/tester')).toBe(path.join('/home/tester', '.webcmd', 'clis'));
         expect(getPluginsDir('/home/tester')).toBe(path.join('/home/tester', '.webcmd', 'plugins'));
     });
-    it('reports unsupported daemon port with WEBCMD env names', () => {
-        const legacyEnvName = ['OPEN', 'CLI_DAEMON_PORT'].join('');
-        expect(unsupportedDaemonPortEnvMessage('1234')).toContain('WEBCMD_DAEMON_PORT');
-        expect(unsupportedDaemonPortEnvMessage('1234')).toContain('Webcmd');
-        expect(unsupportedDaemonPortEnvMessage('1234')).toContain('rerun webcmd');
-        expect(unsupportedDaemonPortEnvMessage('1234')).not.toContain(legacyEnvName);
-    });
 });

package/dist/src/runtime.d.ts

@@ -29,6 +29,7 @@ export interface IBrowserFactory {
         session?: string;
         cdpEndpoint?: string;
         contextId?: string;
+        preferredContextId?: string;
         idleTimeout?: number;
         windowMode?: BrowserWindowMode;
         surface?: BrowserSurface;
@@ -40,6 +41,7 @@ export declare function browserSession<T>(BrowserFactory: new () => IBrowserFact
     session?: string;
     cdpEndpoint?: string;
     contextId?: string;
+    preferredContextId?: string;
     idleTimeout?: number;
     windowMode?: BrowserWindowMode;
     surface?: BrowserSurface;

package/dist/src/runtime.js

@@ -41,6 +41,7 @@ export async function browserSession(BrowserFactory, fn, opts = {}) {
             session: opts.session,
             cdpEndpoint: opts.cdpEndpoint,
             contextId: opts.contextId,
+            preferredContextId: opts.preferredContextId,
             idleTimeout: opts.idleTimeout,
             windowMode: opts.windowMode,
             surface: opts.surface,

package/dist/src/serialization.test.js

@@ -40,22 +40,22 @@ describe('formatRegistryHelpText', () => {
     });
     it('surfaces access and canonical examples instead of strategy as primary help metadata', () => {
         const cmd = {
-            site: 'bilibili',
-            name: 'hot',
+            site: 'youtube',
+            name: 'search',
             access: 'read',
-            description: 'Bilibili hot videos',
+            description: 'Search YouTube',
             strategy: Strategy.COOKIE,
             browser: true,
             args: [],
         };
-        expect(formatCommandExample(cmd)).toBe('webcmd bilibili hot -f yaml');
+        expect(formatCommandExample(cmd)).toBe('webcmd youtube search -f yaml');
         expect(serializeCommand(cmd)).toMatchObject({
-            command: 'bilibili/hot',
+            command: 'youtube/search',
             access: 'read',
-            example: 'webcmd bilibili hot -f yaml',
+            example: 'webcmd youtube search -f yaml',
         });
         expect(formatRegistryHelpText(cmd)).toContain('Access: read');
-        expect(formatRegistryHelpText(cmd)).toContain('Example: webcmd bilibili hot -f yaml');
+        expect(formatRegistryHelpText(cmd)).toContain('Example: webcmd youtube search -f yaml');
         expect(formatRegistryHelpText(cmd)).not.toContain('Strategy:');
     });
     it('surfaces command default output format in structured serialization and help text', () => {

package/dist/src/snapshotFormatter.js

@@ -35,17 +35,17 @@ const CONTAINER_ROLES = new Set([
     'search', 'article', 'paragraph', 'figure',
 ]);
 // Decorator / separator text that adds no semantic value
-const DECORATOR_TEXT = new Set(['•', '·', '|', '—', '-', '/', '\\']);
+const DECORATOR_TEXT = new Set(['•', '\u00b7', '|', '\u2014', '-', '/', '\\']);
 // Ad-related URL patterns
 const AD_URL_PATTERNS = [
     'googleadservices.com/pagead/',
     'alb.reddit.com/cr?',
     'doubleclick.net/',
-    'cm.bilibili.com/cm/api/fees/',
+    'ads.example.com/track/',
 ];
 // Boilerplate button labels to filter (back-to-top, etc.)
 const BOILERPLATE_LABELS = [
-    '回到顶部', 'back to top', 'scroll to top', 'go to top',
+    'back to top', 'scroll to top', 'go to top',
 ];
 /**
  * Parse role and text from a trimmed snapshot line.
@@ -131,8 +131,6 @@ function isAdNode(text, trailingText) {
     const t = (text + ' ' + trailingText).toLowerCase();
     if (t.includes('sponsored') || t.includes('advertisement'))
         return true;
-    if (t.includes('广告'))
-        return true;
     // Check for ad tracking URLs in the label
     for (const pattern of AD_URL_PATTERNS) {
         if (text.includes(pattern) || trailingText.includes(pattern))

package/dist/src/snapshotFormatter.test.d.ts

@@ -1,7 +1,7 @@
 /**
  * Tests for snapshotFormatter.ts: snapshot tree filtering.
  *
- * Uses sanitized excerpts from real websites (GitHub, Bilibili, Twitter)
+ * Uses sanitized excerpts from real websites (GitHub, video sites, Twitter)
  * to validate noise filtering, annotation stripping, and output quality.
  */
 export {};

package/dist/src/snapshotFormatter.test.js

@@ -1,7 +1,7 @@
 /**
  * Tests for snapshotFormatter.ts: snapshot tree filtering.
  *
- * Uses sanitized excerpts from real websites (GitHub, Bilibili, Twitter)
+ * Uses sanitized excerpts from real websites (GitHub, video sites, Twitter)
  * to validate noise filtering, annotation stripping, and output quality.
  */
 import { describe, it, expect } from 'vitest';
@@ -69,54 +69,54 @@ const GITHUB_REPOS = `\
             - img "Repository" [ref=e98]
           - link "agentrhq/webcmd" [ref=e100] [cursor=pointer]:
             - /url: /agentrhq/webcmd`;
-/** Bilibili nav bar (Chinese text, multiple link categories) */
-const BILIBILI_NAV = `\
+/** Video-site nav bar (multiple link categories) */
+const VIDEO_SITE_NAV = `\
 - generic [ref=e3]:
   - generic [ref=e4]:
     - generic [ref=e5]:
       - list [ref=e6]:
         - listitem [ref=e7]:
-          - link "首页" [ref=e8] [cursor=pointer]:
-            - /url: //www.bilibili.com
+          - link "Home" [ref=e8] [cursor=pointer]:
+            - /url: https://video.example.com
             - img [ref=e9]
-            - generic [ref=e11]: 首页
+            - generic [ref=e11]: Home
         - listitem [ref=e12]:
-          - link "番剧" [ref=e13] [cursor=pointer]:
-            - /url: //www.bilibili.com/anime/
+          - link "Shows" [ref=e13] [cursor=pointer]:
+            - /url: https://video.example.com/shows/
         - listitem [ref=e14]:
-          - link "直播" [ref=e15] [cursor=pointer]:
-            - /url: //live.bilibili.com
+          - link "Live" [ref=e15] [cursor=pointer]:
+            - /url: https://live.video.example.com
       - generic [ref=e32]:
-        - textbox "冷知识 金廷26年胜率100%" [ref=e34]
+        - textbox "Search videos" [ref=e34]
         - img [ref=e36] [cursor=pointer]`;
-/** Bilibili video card (deeply nested generic wrappers, view counts) */
-const BILIBILI_VIDEO = `\
+/** Video card (deeply nested generic wrappers, view counts) */
+const VIDEO_CARD = `\
 - generic [ref=e363]:
-  - link "超酷时刻 即将到来 3.3万 40 16:24" [ref=e364] [cursor=pointer]:
-    - /url: https://www.bilibili.com/video/BV1zVw5zoEFt
+  - link "Amazing moment coming soon 33K 40 16:24" [ref=e364] [cursor=pointer]:
+    - /url: https://video.example.com/watch/abc
     - generic [ref=e365]:
-      - img "超酷时刻 即将到来" [ref=e368]
+      - img "Amazing moment coming soon" [ref=e368]
       - generic:
         - generic:
           - generic:
             - generic:
               - img
-              - generic: 3.3万
+              - generic: 33K
             - generic:
               - img
               - generic: "40"
           - generic: 16:24
   - generic [ref=e370]:
-    - heading "超酷时刻 即将到来" [level=3] [ref=e371]:
-      - link "超酷时刻 即将到来" [ref=e372] [cursor=pointer]:
-        - /url: https://www.bilibili.com/video/BV1zVw5zoEFt
-    - link "Tesla特斯拉中国 · 13小时前" [ref=e374] [cursor=pointer]:
-      - /url: //space.bilibili.com/491190876
+    - heading "Amazing moment coming soon" [level=3] [ref=e371]:
+      - link "Amazing moment coming soon" [ref=e372] [cursor=pointer]:
+        - /url: https://video.example.com/watch/abc
+    - link "Global Tech - 13 hours ago" [ref=e374] [cursor=pointer]:
+      - /url: https://video.example.com/channel/global-tech
       - img [ref=e375]
-      - generic "Tesla特斯拉中国" [ref=e379]
-      - generic [ref=e380]: · 13小时前`;
-/** Empty paragraph blocks (Bilibili bottom section) */
-const BILIBILI_EMPTY = `\
+      - generic "Global Tech" [ref=e379]
+      - generic [ref=e380]: 13 hours ago`;
+/** Empty paragraph blocks (video-site bottom section) */
+const VIDEO_SITE_EMPTY = `\
 - generic [ref=e576]:
   - generic:
     - generic:
@@ -354,25 +354,25 @@ describe('formatSnapshot', () => {
             expect(result).not.toContain('/url:');
         });
     });
-    describe('Bilibili snapshot', () => {
-        it('cleans nav bar with Chinese text', () => {
-            const result = formatSnapshot(BILIBILI_NAV);
-            expect(result).toContain('link "首页"');
-            expect(result).toContain('link "番剧"');
-            expect(result).toContain('link "直播"');
-            expect(result).toContain('textbox "冷知识 金廷26年胜率100%"');
+    describe('video-site snapshot', () => {
+        it('cleans nav bar with category links', () => {
+            const result = formatSnapshot(VIDEO_SITE_NAV);
+            expect(result).toContain('link "Home"');
+            expect(result).toContain('link "Shows"');
+            expect(result).toContain('link "Live"');
+            expect(result).toContain('textbox "Search videos"');
             expect(result).not.toContain('[ref=');
         });
         it('handles video card with deeply nested wrappers', () => {
-            const result = formatSnapshot(BILIBILI_VIDEO);
-            expect(result).toContain('link "超酷时刻 即将到来 3.3万 40 16:24"');
-            expect(result).toContain('heading "超酷时刻 即将到来"');
-            expect(result).toContain('generic "Tesla特斯拉中国"');
+            const result = formatSnapshot(VIDEO_CARD);
+            expect(result).toContain('link "Amazing moment coming soon 33K 40 16:24"');
+            expect(result).toContain('heading "Amazing moment coming soon"');
+            expect(result).toContain('generic "Global Tech"');
             // Deeply nested view count generics with text are kept
-            expect(result).toContain('3.3万');
+            expect(result).toContain('33K');
         });
         it('prunes empty paragraph blocks', () => {
-            const result = formatSnapshot(BILIBILI_EMPTY);
+            const result = formatSnapshot(VIDEO_SITE_EMPTY);
             // All content is generic (no text) and empty paragraphs
             // After noise filtering, everything should be pruned
             expect(result.trim()).toBe('');
@@ -424,9 +424,9 @@ describe('formatSnapshot', () => {
             // Expect at least 40% reduction
             expect(formattedLines).toBeLessThan(rawLines * 0.6);
         });
-        it('achieves significant reduction on Bilibili video card', () => {
-            const rawLines = BILIBILI_VIDEO.split('\n').length;
-            const formatted = formatSnapshot(BILIBILI_VIDEO);
+        it('achieves significant reduction on video card', () => {
+            const rawLines = VIDEO_CARD.split('\n').length;
+            const formatted = formatSnapshot(VIDEO_CARD);
             const formattedLines = formatted.split('\n').filter(l => l.trim()).length;
             // Expect at least 30% reduction
             expect(formattedLines).toBeLessThan(rawLines * 0.7);
@@ -463,21 +463,21 @@ describe('formatSnapshot', () => {
             expect(result).toContain('banner "Global Navigation Menu"');
             expect(result).toContain('heading "Dashboard"');
         });
-        it('Bilibili: significant reduction and Chinese text preserved', () => {
-            const raw = loadFixture('snapshot_bilibili.txt');
+        it('video site: significant reduction and navigation text preserved', () => {
+            const raw = loadFixture('snapshot_video.txt');
             if (!raw)
                 return;
             const rawLines = raw.split('\n').length;
             const result = formatSnapshot(raw);
             const resultLines = result.split('\n').filter((l) => l.trim()).length;
-            // Should achieve > 40% reduction on Bilibili (lots of imgs and generics)
+            // Should achieve > 40% reduction on a video-site snapshot (lots of imgs and generics)
             expect(resultLines).toBeLessThan(rawLines * 0.6);
             // No annotations remain
             expect(result).not.toContain('[ref=');
             expect(result).not.toContain('[cursor=');
-            // Chinese text preserved
-            expect(result).toContain('link "首页"');
-            expect(result).toContain('link "番剧"');
+            // Navigation text preserved
+            expect(result).toContain('link "Home"');
+            expect(result).toContain('link "Shows"');
         });
         it('Twitter/X: significant reduction and tweet structure preserved', () => {
             const raw = loadFixture('snapshot_twitter.txt');
@@ -508,8 +508,8 @@ describe('formatSnapshot', () => {
             expect(result).toContain('link');
             expect(result).toContain('textbox');
         });
-        it('Bilibili maxDepth=3: shallow view', () => {
-            const raw = loadFixture('snapshot_bilibili.txt');
+        it('video site maxDepth=3: shallow view', () => {
+            const raw = loadFixture('snapshot_video.txt');
             if (!raw)
                 return;
             const result = formatSnapshot(raw, { maxDepth: 3 });

package/dist/src/tui.js

@@ -37,7 +37,7 @@ export async function checkboxPrompt(items, opts = {}) {
             const status = item.status ?? '';
             out += `  ${pointer} ${checkbox} ${item.label}${status ? `  ${status}` : ''}\n`;
         }
-        out += `\n  ↑↓ navigate  ·  Space toggle  ·  a all  ·  Enter confirm  ·  q cancel\n`;
+        out += `\n  Up/Down navigate | Space toggle | a all | Enter confirm | q cancel\n`;
         return out;
     }
     return new Promise((resolve) => {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@agentrhq/webcmd",
-  "version": "0.1.2",
+  "version": "0.2.1",
   "description": "Turn websites, browser sessions, desktop apps, and local tools into deterministic CLI surfaces for humans and AI agents.",
   "engines": {
     "node": ">=20.0.0"
@@ -45,9 +45,10 @@
     "build-manifest": "tsx src/build-manifest.ts",
     "clean-dist": "node scripts/clean-dist.cjs",
     "copy-yaml": "node scripts/copy-yaml.cjs",
+    "generate-release-notes": "tsx scripts/generate-release-notes.ts",
     "start": "node dist/src/main.js",
     "start:bun": "bun dist/src/main.js",
-    "preuninstall": "node -e \"fetch('http://127.0.0.1:19825/shutdown',{method:'POST',headers:{'X-Webcmd':'1'},signal:AbortSignal.timeout(3000)}).catch(()=>{})\" || true",
+    "preuninstall": "node -e \"fetch('http://127.0.0.1:9777/shutdown',{method:'POST',headers:{'X-Webcmd':'1'},signal:AbortSignal.timeout(3000)}).catch(()=>{})\" || true",
     "postinstall": "node scripts/postinstall.js || true; node scripts/fetch-adapters.js || true",
     "typecheck": "tsc --noEmit",
     "prepare": "[ -d src ] && npm run build || true",
@@ -90,6 +91,8 @@
     "ws": "^8.18.0"
   },
   "devDependencies": {
+    "@emnapi/runtime": "^1.11.2",
+    "@google/genai": "^2.10.0",
     "@types/js-yaml": "^4.0.9",
     "@types/jsdom": "^27.0.0",
     "@types/node": "^25.5.2",