What if we do something completely radical? Instead of bickering endlessly over NASA's budget every year, let's assign them a set of priorities (much like we already do) and then fund them for the next ten years. Set aside $225 billion (covering the current budget level plus a hefty increase for expanded programs, growing at 4% per year for the next 10 years) in an interest-bearing account. Now there is no budget uncertainty at all and NASA can do longer-term planning without risking a funding cut halfway through an expensive project. This will save money in the long run since less will be wasted on pivots or on research into projects that get canceled. If the agency saves money in a given year they will have more money available in a later year automatically.
10-year treasury bonds are going for about 1.5% right now, for an overall cost of $261.12 billion or total financing costs of $36.12 billion. We would have the option of paying or not paying into that debt over the next ten years, then the option of paying it off or rolling it into new bonds. We would also get to decide at that point whether to do another 10-year funding run or go back to traditional budgeting.
Obviously accountability is a concern, so there would still be a need for an annual budget, annual review and a report to Congress on activities and progress so far. On the other hand, congresscritters and Presidents won't be able to pull course-changing publicity stunts or divert funding to their pet companies. As a check against too much individual power, the administrator could still be fired and replaced if they are running off the rails.