The best solution is to at first do nothing and check all the remaining parameters
I like that answer! I observe too often that in cruise flight something will appear to "happen" or change, and a pilot thinks that they need to apply their cat like reflexes to compensate for it. Sometimes it's better just to have already been aware, and then continue to be aware, perhaps with an added element to consider. As time passes, and other factors require a change, then maybe compensation, or a changed plan is going to be needed. Maybe, nothing (other than a written up snag) will be required.
Which reminds me of one of my learning events. I was first time left seat ferrying a Twin Otter with a very experiences mentor pilot friend. We were leaving Cairo southbound, and low altitude nav aids were few and far between. This particular Twin Otter did not have a DG at all, it had two slaved RMI's and the magnetic compass (which is not remarkably useful for flying a heading in a Twin Otter). We had noticed on previous legs that the RMI's would drop a flag an quit, seemingly randomly, so becoming useless for flying a heading. We agreed that whomever had a working RMI would fly (no auto pilot, hand flying the whole trip anyway). So, when a half hour into a 9 hour leg, my RMI dropped a flag, I wasn't really eager to surrender flying just yet, I was enjoying myself! My mentor friend was consumed with a marine "Sat Nav" device he had brought on the trip. This was before the days of GPS - the time before magenta lines, so it was charts and heading indicators, other than Bill's occasional fixes on his Sat Nav. He was happy watching the Sat Nav, and probably had little interest in flying anyway, so I kept flying... but how was I going to hold a heading? I could look across at his RMI, but that was not a really good long term solution. As it was a very clear day, I could fly by ground reference for a while, but in that part of the world, sand is sand, so there were not many features to pick on the horizon.
So I flew on, taking my time, and thinking. During this period, I found a solution. I considered it, verified it with occasional glances to Bill's RMI, and applied it. I flew a perfect track. Eventually, Bill noticed that my RMI had dropped a flag, he commented without alarm. He asked if I was okay continuing flying, and I said I was. Over the next number of hours, he used the Sat Nav to confirm that my track was right on, and eventually asked me how I was doing it. I told him I'd tell him later.
25 flying hours later, (all but the final leg of which I was offered left seat), and many more RMI failures, I had kept my technique to myself, and he seemed content to ride along, confirming my track, and perfecting his use of the Sat nav for sailing (which interested him more than flying a Twin Otter). We arrived in Maseru and delivered the plane to Air Lesotho. At dinner he finally said: "Okay, you gotta tell me how you were doing that, you were flying perfect tracks for hours with no practical heading indicator." I explained that while flying with the RMI card stopped, I noticed that the RMI slaving meter would show the failing attempts of the remote compass to slave the card. As long as the RMI failed on the heading I wanted to fly, and I kept the slaving meter centered, to plane followed the heading. He quietly smiled. I learned to determine if something was a problem, before doing something to solve it.