If the task is unfamiliar and time is limited (for example, a week), how do you act?

When faced with an unfamiliar task and limited time, I usually proceed as follows:

1. Quick analysis of the task — understand exactly what needs to be done, what requirements and constraints exist.
2. Search and study ready-made solutions — look for libraries, tools, or approaches that can be used to avoid reinventing the wheel.
3. Break down the task into subtasks — identify the minimal necessary steps to achieve the result.
4. Prioritization — determine what must be done first and what can be postponed or simplified.
5. Prototyping — create a minimally working version to test key hypotheses.
6. Iterative improvement — gradually add functionality and improve quality.

It is important not to try to do everything perfectly right away, but to focus on achieving a working result within the deadline. If questions arise, I am not shy to consult colleagues or look for information in documentation and communities.