Files
BrowserOS/apps/server/tests/tools/controller-based/content.test.ts
Nikhil afddda015a feat: fix imports to remove .js (#170)
* fix: remove all .js in imports

* fix: update claude mode to use right import

* fix: remove addition in main package.json
2026-01-06 10:54:26 -08:00

510 lines
18 KiB
TypeScript

// @ts-nocheck
/**
* @license
* Copyright 2025 BrowserOS
*/
import { describe, it } from 'bun:test'
import assert from 'node:assert'
import { withMcpServer } from '../../__helpers__/utils'
describe('MCP Controller Content Tools', () => {
describe('browser_get_page_content - Success Cases', () => {
it('tests that page content extraction with text type succeeds', async () => {
await withMcpServer(async (client) => {
// Navigate to a page with content
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><h1>Title</h1><p>This is a paragraph of text.</p><p>Another paragraph.</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text' },
})
console.log('\n=== Get Page Content (Text) Response ===')
console.log(JSON.stringify(result, null, 2))
assert.ok(Array.isArray(result.content), 'Content should be array')
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(textContent, 'Should have text content')
// If getSnapshot API is available, check for pagination info
if (!result.isError && textContent.text.includes('Total pages:')) {
assert.ok(
textContent.text.includes('characters total'),
'Should include character count',
)
}
})
}, 30000)
it('tests that page content extraction with text-with-links type succeeds', async () => {
await withMcpServer(async (client) => {
// Navigate to a page with links
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><h1>Links Page</h1><a href="https://example.com">Example Link</a><p>Some text</p><a href="https://test.com">Test Link</a></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text-with-links' },
})
console.log('\n=== Get Page Content (Text with Links) Response ===')
console.log(JSON.stringify(result, null, 2))
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(textContent, 'Should have text content')
// If getSnapshot API is available, check for pagination info
if (!result.isError) {
assert.ok(
textContent.text.includes('Total pages:') ||
textContent.text.includes('Error:'),
'Should include pagination info or error',
)
}
})
}, 30000)
it('tests that page content extraction with specific page number succeeds', async () => {
await withMcpServer(async (client) => {
// Navigate to a page with content
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><h1>Page Title</h1><p>Content here</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text', page: '1' },
})
console.log('\n=== Get Page Content (Page 1) Response ===')
console.log(JSON.stringify(result, null, 2))
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(textContent, 'Should have text content')
// If getSnapshot API is available, check for page info
if (!result.isError) {
assert.ok(
textContent.text.includes('Page 1 of') ||
textContent.text.includes('Error:'),
'Should indicate page 1 or error',
)
}
})
}, 30000)
it('tests that page content extraction with all pages succeeds', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><h1>Title</h1><p>Content</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text', page: 'all' },
})
console.log('\n=== Get Page Content (All Pages) Response ===')
console.log(JSON.stringify(result, null, 2))
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(textContent, 'Should have text content')
// If getSnapshot API is available, check for total pages
if (!result.isError) {
assert.ok(
textContent.text.includes('Total pages:') ||
textContent.text.includes('Error:'),
'Should show total pages or error',
)
}
})
}, 30000)
it('tests that page content extraction with different context window sizes succeeds', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><h1>Title</h1><p>Content</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
// Test different context windows
const contextWindows = ['20k', '30k', '50k', '100k']
for (const contextWindow of contextWindows) {
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text', contextWindow },
})
console.log(
`\n=== Get Page Content (${contextWindow} window) Response ===`,
)
console.log(JSON.stringify(result, null, 2))
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(textContent, 'Should have text content')
// If getSnapshot API is available, check for context window info
if (!result.isError) {
assert.ok(
textContent.text.includes(contextWindow) ||
textContent.text.includes('Error:'),
`Should mention ${contextWindow} or error`,
)
}
}
})
}, 60000)
it('tests that empty page content extraction is handled', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text' },
})
console.log('\n=== Get Page Content (Empty Page) Response ===')
console.log(JSON.stringify(result, null, 2))
assert.ok(!result.isError, 'Should succeed')
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(textContent, 'Should have text content')
})
}, 30000)
})
describe('browser_get_page_content - Error Handling', () => {
it('tests that content extraction with invalid tab ID is handled', async () => {
await withMcpServer(async (client) => {
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId: 999999999, type: 'text' },
})
console.log('\n=== Get Page Content Invalid Tab Response ===')
console.log(JSON.stringify(result, null, 2))
assert.ok(result, 'Should return a result')
assert.ok(Array.isArray(result.content), 'Should have content array')
if (result.isError) {
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(textContent, 'Error should include text content')
}
})
}, 30000)
it('tests that non-numeric tab ID is rejected', async () => {
await withMcpServer(async (client) => {
try {
await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId: 'invalid', type: 'text' },
})
assert.fail('Should have thrown validation error')
} catch (error) {
console.log('\n=== Get Page Content Invalid Tab Type Error ===')
console.log(error.message)
assert.ok(
error.message.includes('Invalid arguments') ||
error.message.includes('Expected number'),
'Should reject with validation error',
)
}
})
}, 30000)
it('tests that invalid type enum is rejected', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><p>Content</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
try {
await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'invalid-type' },
})
assert.fail('Should have thrown validation error')
} catch (error) {
console.log('\n=== Get Page Content Invalid Type Error ===')
console.log(error.message)
assert.ok(
error.message.includes('Invalid') || error.message.includes('enum'),
'Should reject with validation error',
)
}
})
}, 30000)
it('tests that invalid page number is handled', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><p>Content</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text', page: '999' },
})
console.log('\n=== Get Page Content Invalid Page Response ===')
console.log(JSON.stringify(result, null, 2))
assert.ok(!result.isError, 'Should not throw error')
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(
textContent.text.includes('Error') ||
textContent.text.includes('Invalid page'),
'Should indicate invalid page',
)
})
}, 30000)
it('tests that non-numeric page number is handled', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><p>Content</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text', page: 'invalid' },
})
console.log('\n=== Get Page Content Non-Numeric Page Response ===')
console.log(JSON.stringify(result, null, 2))
assert.ok(!result.isError, 'Should not throw error')
const textContent = result.content.find((c) => c.type === 'text')
assert.ok(
textContent.text.includes('Error') ||
textContent.text.includes('Invalid page'),
'Should indicate invalid page',
)
})
}, 30000)
})
describe('browser_get_page_content - Response Structure Validation', () => {
it('tests that content tool returns valid MCP response structure', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><h1>Test</h1><p>Content</p></body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
const result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text' },
})
// Validate response structure
assert.ok(result, 'Result should exist')
assert.ok('content' in result, 'Should have content field')
assert.ok(Array.isArray(result.content), 'content must be an array')
if ('isError' in result) {
assert.strictEqual(
typeof result.isError,
'boolean',
'isError must be boolean when present',
)
}
// Validate content items
for (const item of result.content) {
assert.ok(item.type, 'Content item must have type')
assert.ok(
item.type === 'text' || item.type === 'image',
'Content type must be text or image',
)
if (item.type === 'text') {
assert.ok('text' in item, 'Text content must have text property')
assert.strictEqual(
typeof item.text,
'string',
'Text must be string',
)
}
}
})
}, 30000)
})
describe('browser_get_page_content - Workflow Tests', () => {
it('tests complete content extraction workflow: navigate -> extract text -> extract text-with-links', async () => {
await withMcpServer(async (client) => {
// Navigate to a page
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url: 'data:text/html,<body><h1>Article Title</h1><p>This is a paragraph with <a href="https://example.com">a link</a>.</p><h2>Subtitle</h2><p>More content here.</p></body>',
},
})
console.log('\n=== Workflow: Navigate Response ===')
console.log(JSON.stringify(navResult, null, 2))
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
// Extract text only
const textResult = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text' },
})
console.log('\n=== Workflow: Extract Text ===')
console.log(JSON.stringify(textResult, null, 2))
assert.ok(!textResult.isError, 'Text extraction should succeed')
// Extract text with links
const linksResult = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text-with-links' },
})
console.log('\n=== Workflow: Extract Text with Links ===')
console.log(JSON.stringify(linksResult, null, 2))
assert.ok(
!linksResult.isError,
'Text with links extraction should succeed',
)
})
}, 30000)
it('tests pagination workflow: extract all pages -> extract specific page', async () => {
await withMcpServer(async (client) => {
const navResult = await client.callTool({
name: 'browser_navigate',
arguments: {
url:
'data:text/html,<body><h1>Long Content</h1><p>'.repeat(100) +
'Content paragraph.' +
'</p>'.repeat(100) +
'</body>',
},
})
const navText = navResult.content.find((c) => c.type === 'text')
const tabIdMatch = navText.text.match(/Tab ID: (\d+)/)
const tabId = parseInt(tabIdMatch[1], 10)
// Extract all pages with small context window
const allPagesResult = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text', page: 'all', contextWindow: '20k' },
})
console.log('\n=== Workflow: Extract All Pages ===')
console.log(JSON.stringify(allPagesResult, null, 2))
assert.ok(
!allPagesResult.isError,
'All pages extraction should succeed',
)
// Extract specific page
const page1Result = await client.callTool({
name: 'browser_get_page_content',
arguments: { tabId, type: 'text', page: '1', contextWindow: '20k' },
})
console.log('\n=== Workflow: Extract Page 1 ===')
console.log(JSON.stringify(page1Result, null, 2))
assert.ok(!page1Result.isError, 'Page 1 extraction should succeed')
})
}, 30000)
})
})