Loading...
{
"mcpServers": {
"scrapelessMcpServer": {
"command": "npx",
"args": ["-y", "scrapeless-mcp-server"],
"env": {
"SCRAPELESS_KEY": "YOUR_SCRAPELESS_KEY"
}
}
}
}google_searchQueries Google Search and returns search results.
Read-only operation; no modification of data.
google_trendsRetrieves trending search data from Google Trends.
Read-only operation; no modification of data.
browser_createCreates or reuses a cloud browser session.
Creates a browser session that can be used for potentially harmful actions.
browser_closeCloses the current browser session.
Terminates a session, preventing further actions.
browser_gotoNavigates the browser to a specified URL.
Navigates to arbitrary URLs, which could be malicious.
browser_go_backGoes back one step in the browser history.
Navigates browser history; no external impact.
browser_go_forwardGoes forward one step in browser history.
Navigates browser history; no external impact.
browser_clickClicks a specific element on the page.
Can trigger actions on a webpage.
browser_typeTypes text into a specified input field.
Can input data into forms, potentially submitting sensitive information.
browser_press_keySimulates a key press in the browser.
Can trigger actions via keyboard shortcuts.
browser_wait_forWaits for a specific page element to appear.
Passive observation; no impact.
browser_waitPauses execution for a fixed duration.
Passive delay; no impact.
browser_screenshotCaptures a screenshot of the current page.
Captures an image of the current state; no external impact.
browser_get_htmlGets the full HTML of the current page.
Extracts content; no external impact.
browser_get_textGets all visible text from the current page.
Extracts content; no external impact.
browser_scrollScrolls to the bottom of the page.
Navigates the page; no external impact.
browser_scroll_toScrolls a specific element into view.
Navigates the page; no external impact.
scrape_htmlScrapes a URL and returns its full HTML content.
Extracts content; no external impact.
scrape_markdownScrapes a URL and returns its content as Markdown.
Extracts content; no external impact.
scrape_screenshotCaptures a high-quality screenshot of any webpage.
Captures an image of the current state; no external impact.
The Scrapeless MCP Server is relatively safe for read-only operations like search and data extraction from trusted sources. However, using browser automation and scraping untrusted sites introduces significant risks if the data isn't properly sanitized and validated before being used in LLM prompts or other applications.