Npilot
AI-generatedSummary
Execute a series of browser automation steps within a managed session, allowing control over existing or new sessions and supporting scripted interactions such as navigation, clicks, typing, extraction, and more.
Inputs
- Session (required) — Specify whether to open a new browser session or reuse a previous session by providing its Session ID.
- Session ID — Provide the Session ID of an existing browser session to reuse; required if using previous session mode.
- Operation Modes (required) — Choose between 'Action Based' mode for configuring discrete steps in the UI or 'Script Based (JSON)' mode to provide a JSON script of steps.
- Steps (Action Based) — Define an ordered list of steps each specifying an action (e.g., navigate, click, type) with associated parameters like selectors, URLs, text, timeouts, and output field names.
- Script Based (JSON) — Provide a JSON array of step objects where each object defines an action key and its parameters, executed in order for dynamic scripting scenarios.
- Additional Options — Configure browser session options such as extra HTTP headers, headless mode, TTL, user agent, and viewport size.
Output shape
a list of items corresponding to input items, each with updated JSON containing session ID and any step-specific output fields; binary output fields included for screenshot steps.
Each input item produces an output item containing original JSON merged with sessionId and results of steps under user-defined output fields. Screenshot data is stored as binary. Errors can be optionally caught via continueOnFail setting. Session management actions (e.g., closeSession) affect session lifetime.