My #1 tip would be to find out what the _real_ success criteria (and metrics for accessing them) for the project are (hint: it’s never hitting milestones, or coming in on budget) and focus on meeting them.
Forget the triangle talk; you can deliver a successful project that is late, over budget and low quality.
If there genuinely are not measurable success criteria, or they’re long forgotten or no longer relevant; then get the project shit-canned on this basis because otherwise it’s a death-march doomed to fail.