Search & fetch
| Tool | Parameters | Description |
|---|
web_search | query, n? | Search the web and return the top N results (default 5) with title, URL, and snippet. |
fetch_url | url, as_markdown? | Fetch a URL and return the page content. Set as_markdown to get cleaned readable text instead of raw HTML. |
search_images | query, n? | Search for images by query. Returns URLs and metadata for matching image results. |
Browser automation
These tools drive a headless Chromium browser for interactive web tasks — filling forms, clicking buttons, scraping dynamic content, and taking screenshots.
| Tool | Parameters | Description |
|---|
browse_url | url | Open a URL in the headless browser and return the rendered page content as markdown. |
screenshot_page | url, full_page?, output_name? | Take a screenshot of a page and save it as a PNG artifact. Supports full-page capture. |
browser_navigate | url | Navigate the browser to a new URL within an active browser session. |
browser_click | selector | Click an element identified by a CSS selector. |
browser_type | selector, text | Type text into an input field identified by a CSS selector. |
browser_fill | fields | Fill multiple form fields at once. Pass an object mapping selectors to values. |
browser_extract | selector, attribute? | Extract text or an attribute value from elements matching a CSS selector. |
HTTP requests
| Tool | Parameters | Description |
|---|
http_get | url, headers? | Make an HTTP GET request and return the response body and status code. |
http_request | method, url, headers?, body? | Make any HTTP request (GET, POST, PUT, DELETE, etc.) with optional headers and body. Returns status, headers, and response body. |
