I do not only just “my part of work”, but always look how it will be integrated with other things.
Work is done, when it is accepted by stakeholders, and they receive certain value from it. No point in continuing doing things, if they are not going to be accepted.
Defining what will be tested and how often bring up hidden requirements and constraints.
I always try to find the root problem, even if stakeholders are unaware of it. Solving wrong things is often a huge waste of time.
If some process doesn't work, I try to discuss if it is still required.
I ask questions. I ask why things are done in a certain way until others start feeling uncomfortable. It is ok to have reasons such as: personal preference, conscious decision, legal constraints, availability of skills in the team, nobody objected, wrong implementation, reducing time to market, experimenting with new technology.
The world is changing constantly, and old ways of working may become inefficient or even harmful.
Linux over Windows over Mac. Lenovo over other laptop brands.
I try to create a safe space for experimenting and adopting new technologies in the team. By creating clear guidelines on how to integrate new tools and processes and clear guidelines on how to stop using them we reduce obsession with the legacy.