Actions108
- AI Visibility Actions
- List Sources
- Search
- Agentic Search
- AI Visibility Search
- AI Visibility Sources
- Browser Task
- Control Monitor
- Crawl
- Create Monitor
- Delete Session
- Get Monitor Changes
- List Monitors
- List Sessions
- Map
- Scrape URL
- Wire: Browse Catalog
- Wire: Discover Actions
- Wire: List Identities
- Wire: Request New Action
- Wire: Run Read Action
- Wire: Run Write Action
- Wire: Sign In
- Browser Actions
- Delete Session
- List Sessions
- Run Task
- Agentic Search
- AI Visibility Search
- AI Visibility Sources
- Browser Task
- Control Monitor
- Crawl
- Create Monitor
- Get Monitor Changes
- List Monitors
- Map
- Scrape URL
- Search
- Wire: Browse Catalog
- Wire: Discover Actions
- Wire: List Identities
- Wire: Request New Action
- Wire: Run Read Action
- Wire: Run Write Action
- Wire: Sign In
- Monitor Actions
- Control
- Create
- Get Changes
- List
- Agentic Search
- AI Visibility Search
- AI Visibility Sources
- Browser Task
- Control Monitor
- Crawl
- Create Monitor
- Delete Session
- Get Monitor Changes
- List Monitors
- List Sessions
- Map
- Scrape URL
- Search
- Wire: Browse Catalog
- Wire: Discover Actions
- Wire: List Identities
- Wire: Request New Action
- Wire: Run Read Action
- Wire: Run Write Action
- Wire: Sign In
- Web Actions
- Agentic Search
- Crawl
- Map
- Scrape URL
- Search
- AI Visibility Search
- AI Visibility Sources
- Browser Task
- Control Monitor
- Create Monitor
- Delete Session
- Get Monitor Changes
- List Monitors
- List Sessions
- Wire: Browse Catalog
- Wire: Discover Actions
- Wire: List Identities
- Wire: Request New Action
- Wire: Run Read Action
- Wire: Run Write Action
- Wire: Sign In
- Wire Actions
Browser → Run Task
AI-generatedSummary
Run a natural-language browser automation task using a real AI-driven browser to perform actions such as navigation, clicking, typing, and data extraction.
Inputs
- Prompt (required) — The natural language task description specifying the goal and expected result. Be specific, and do not include passwords or secrets (use Session ID for authentication).
- Additional Options — Optional settings to control task execution, including:
- Max Steps: Maximum number of agent steps (navigation/click/type actions), 0 for default.
- Max Wait Time (Seconds): Maximum time to wait for task completion (default 360s).
- Output Schema (JSON): JSON Schema defining expected structured output.
- Poll Interval (Seconds): Time between status checks (default 3s).
- Session ID: Saved browser session ID to run the task logged in.
- Timeout (Ms): Task timeout in milliseconds; server caps at about 330,000 ms.
- URL: URL to navigate to before starting; leave empty to let the agent follow URLs named in the prompt.
Output shape
a single JSON object representing the completed browser task result
The operation returns immediately after submitting the task, then polls until completion or timeout. The result includes success status, workflow ID, and task data conforming to the specified output schema if provided. On failure or timeout, an error is raised.
Examples
Example 1: Extract product data by describing the task in natural language, optionally specifying a URL to start and an output JSON schema.
Provide the task prompt describing the goal, set optional parameters like max wait time or session ID for authenticated sites, and run the task. The output includes structured data retrieved by the AI-driven browser.