Web Page Content Extraction
This skill guides you on how to extract text and image content from any publicly accessible web page via the LinkFox tool gateway. It launches a remote browser to render the target page, then returns the extracted text, images, and optional file attachments.
Core Concepts
The Web Extract tool uses a remote headless browser to fully render a target web page (including JavaScript-heavy SPAs), then extracts all visible text and images from the rendered result. Because it performs real browser rendering, it can handle dynamic content that simple HTTP fetches cannot.
Performance note: This tool launches a remote browser instance, so it is slower than other tools. Prefer lightweight alternatives when available. Use this tool when the user explicitly provides a URL and wants to read or analyze the page content.
Unstructured output: The extracted content is returned in a non-fixed format. Do not attempt to process the result with sandbox code execution or SQL queries. Analyze and summarize the content directly in your response.
Data Fields
Request Parameters
| Field | Description | Required | Example |
|---|---|---|---|
| url | Target web page URL (must start with http:// or https://) | Yes | https://example.com/article |
| userPrompt | Natural language description of what the user wants from the page | Yes | Extract the main article text and all product images |
| memberId | Member identifier for the requesting user | Yes | user_12345 |
| uid | User ID | No | uid_001 |
| chatId | Conversation ID | No | chat_abc |
| requestId | Push/request ID | No | req_xyz |
| teamId | Team ID | No | team_001 |
Response Fields
| Field | Type | Description |
|---|---|---|
| text | string | Extracted text content from the web page |
| images | array | Array of image objects with title, url, and timestamp |
| files | array | Array of file objects with name, type, content (URL), and size |
| finished | boolean | Whether the extraction task completed |
| status | string | Browser status indicator |
| taskId | string | Unique task identifier |
| title | string | Page title |
| type | string | Type marker (browser rendering) |
| costToken | integer | Tokens consumed by the extraction |
| id | integer | Result ID |
Parameter Guide
Writing the userPrompt
The userPrompt tells the remote browser what to focus on when extracting content. Write it to reflect the user's actual intent.
Examples:
1. General page content extraction
url: https://example.com/blog/post-123
userPrompt: Extract all the text content from this article page
2. Focused image extraction
url: https://example.com/products/widget
userPrompt: Extract all product images and their descriptions from this page
3. Specific data extraction
url: https://example.com/pricing
userPrompt: Extract the pricing table and plan comparison details
4. News or article summarization
url: https://news.example.com/2026/03/story
userPrompt: Extract the full article text including headline, author, and publication date
5. Technical documentation extraction
url: https://docs.example.com/api/reference
userPrompt: Extract the API endpoint documentation, parameters, and code examples
Display Rules
- Present extracted content clearly: Format the extracted text in a readable structure. Use headings, paragraphs, and lists as appropriate based on the content
- Show images: When the response includes images, present them with their titles and URLs so the user can view them
- File attachments: If the response includes files, clearly list the file names and download URLs
- Summarize when requested: If the user asked to summarize or analyze the page, provide your analysis based on the extracted content rather than dumping raw text
- Performance expectation: Inform users that extraction may take a moment since a remote browser is rendering the page
- Error handling: When extraction fails, explain the issue based on the response status and suggest checking the URL accessibility or trying again
API Usage
This tool calls the LinkFox tool gateway API. See references/api.md for calling conventions, request parameters, and response structure. You can also execute scripts/web_extract.py directly to run extractions.
Applicable Scenarios
| User Says | Scenario |
|---|---|
| "What's on this page?" / pastes a URL | General page content extraction |
| "Read this article for me" | Article text extraction |
| "Grab the images from this site" | Image extraction |
| "Summarize this web page" | Content extraction + summarization |
| "Extract the data from this URL" | Structured data extraction |
| "What does this link say?" | Quick page content review |
| "Scrape this webpage" | Full page content extraction |
| "Fetch the content from this URL" | General web content retrieval |
Not Applicable Scenarios
- Local file processing: User wants to process a file already on their machine (not a URL)
- API calls: User wants to call a REST API endpoint directly (use HTTP tools instead)
- PDF reading: User has a PDF file to read (use PDF tools instead)
- Screenshot/rendering: User wants a visual screenshot of a page (this tool extracts text/images, not screenshots)
- Authenticated pages: Pages behind login walls may not be accessible since the remote browser has no stored credentials
- Real-time monitoring: Continuously watching a page for changes is not supported
Boundary judgment: When users paste a URL and want to understand its content, this skill applies. If they want to interact with a web application (fill forms, click buttons, navigate multi-step flows), this tool provides a single-page extraction only and may not capture the full interactive experience.
Feedback:
Auto-detect and report feedback via the Feedback API when any of the following apply:
- The functionality or purpose described in this skill does not match actual behavior
- The skill's results do not match the user's intent
- The user expresses dissatisfaction or praise about this skill
- Anything you believe could be improved
Call the feedback API as specified in references/api.md. Do not interrupt the user's flow.
For more high-quality, professional cross-border e-commerce skills, set LinkFox Skills.