How can one train prophylaxis in Chess?

I noticed that I tend to ignore or underestimate my opponent plan, how can I train and improve my prophylaxis ?
Should I do many tactics daily ? I noticed that I tend to ignore my opponent's plan when I don't really know the opening or the variation that we're playing, so should I study my openings a little bit more ?

Force yourself to consider opponent plans

You can't really "train" for prophylaxis. It's like training for "aggressiveness"...a bit too vague for such a thing.

It's a very general topic, but I think there are ways you can work on it, just like you can work on strategy by breaking it down into smaller elements.

For tactical practice, you could flip the board when solving puzzles, so that you get in the habit of looking for tactics from a defensive point of view. I also think endgame studies are probably good for thinking about defensive ideas in a concrete way.

Prophylaxis usually refers to positional ideas, which are harder to work on, but: (1) You might take a positional handbook and think about each element from a defensive point of view. For example, there is a lot of material about how to use the advantage of the bishop pair, but from a prophylactic standpoint you should realize that one of your opponent's plans is to eliminate your bishop pair, and look for ways to avoid that. (2) There are also positional puzzle books and you could solve those from the other side of the board. (3) You could study games of masters known for prophylaxis like Petrosian and Karpov. (4) If you play longer time controls online, you could flip the board regularly, just to force yourself to think about your opponent's plan. (5) There are some resources you can use. I watched the free lesson from Kramnik's Chessable course Thinking In Chess, and he says that prophylaxis is probably the main idea in chess, so I imagine he talks a lot about it, but I think the course is probably too advanced for me. Jacob Aagaard has a book called Practical Chess Defense which has lots of puzzles like all his books, and Mark Dvoretsky has a book called Recognizing Your Opponent's Resources: Developing Preventative Thinking, which also has lots of puzzles. I have only looked at a few puzzles from these books, however, and they tend to be very hard. It'd be nice if there were more material on this topic for intermediate players, but I imagine it doesn't sell as well.

One piece of advice I found useful was this. Seek to expel, exchange, capture or undermine any piece or pawn on your half of the board. As soon as your opponent gets anything on your half of the board, your alarm bells should ring. Examples are the knight on e5 when you are black against the London System and the advanced pawn in advanced variations of the French or Caro-Kann. Other examples are bishops applying pins.

Of course, you have to apply these principles with tactical analysis. For example, if you seek to nullify the knight on e5 in the London System by capturing it, you might promote the capturing pawn to e5 which then pushes away your king's knight from its key defensive post on f6. Advanced bishops and advanced knights on a knight file must often be repelled with a rook pawn push. Again, apply with proper tactical analysis calculations.

Taking this one step further, prophylaxis will be when you make moves which prevent the enemy pieces or pawns coming to important squares on your side of the board. Equally, you must abstain from certain moves which will only promote, by capture, something even more dangerous to your side of the board.

Then there is the issue of weak squares or holes. Don't push pawns without an excellent tactical / strategic reason or you will leave uncompensated holes in your position. Find lessons on weak squares / holes and assimilate them into your understanding. Prophylaxis here will about avoiding making weakening moves which leave holes in your position.

At the level of pieces, avoid exchanging a piece, often a bishop, which is crucial to covering holes in your position. I had a game where I had a lot of weak dark squares. This often happens when you fianchetto the king's bishop as black and push the e pawn as well. This combination is often a bad idea. I had this pawn structure with the bishop not on the fianchetto square. Bad idea. Then I compounded that mistake by exchanging off that bishop for a miniscule reduction of pressure on me in the short term. Bad idea. The dark square holes / weaknesses were cramping and crippling. I lost.

Final note, while worrying about advanced enemy pieces and pawns and not creating holes in your position, don't forget about long range pieces lurking way back in the enemy position. Fianchettoed bishops are a prime example but not the only example. Always draw a mental highlighter line from them to all the squares they can potentially reach on your side of the board. Draw this mental line irrespective of intervening pieces, his or yours, to see what it could threaten. In many positions, diagonals and straight lines can be cleared surprisingly fast by tactical combinations and you don't want to suffer a nasty discovered attack at the end of a combination.

Hope this helps. I am still early in my chess journey. Knowing these principles is one thing. Being able to remember them all, see all the possibilities and apply them in practical play under time limits is very hard. I am still working on it myself so I don't claim to be any kind of expert.

whenever you feel a danger, consider your opponent best move..

As you get stronger you will consider your opponents ideas - maybe even more than your own. That's just what it takes to be a strong player. The gap is generally in visualization of a position not on the board (but in your minds eye). Because that skill is lacking, or because time controls are too fast... we gloss over deep positions as players.

It is possible to play a reasonably good game without watching your opponents ideas deeply - but you will be surprised and lose if they should have a good idea. The advice is always the same for this, yet few ever take it: Stop playing fast chess and focus on your opponents stuff until you figure it out. Under 30 minutes per side is usually too fast to learn this skill. Visualize deeply positions not on the board. Play blindfolded if you have to - but it is not necessary to do so to improve this.

I think I made the biggest leaps forward by doing composed problems and studies as they often require deep analysis from the initial position.

Warm up with 6 training games using 90% defense. Just don't stay in hedgehog waiting to get smothered. still slowly advance but always check for weak pawns and covering squares to keep opponents pieces on their side of board. Target most active threatening piece. When advancing dont leave things unprotected left behind.

Then do 3 training games following strategy for your opening while keeping those defensive measures in mind. If those training games go well take a break, eat, hydrate, then play 2 or 3 rated games.

Go over some games of Tigran Petrosian (extreme) or Karpov or Kramnik
The free site playgrandmasters.com
Has game collection for these and other players where you can do Find Best Move approach

I tried to organize these based on suitability for training. You can use a FindBestMove approach with games so that you have to be engaged and work hard to find good moves.

1. Mastering Chess Strategy, Hellsten Ch5 Prophylaxis about 30 pp., plus 60 related Ch 5 positions in the separate Exercises Section to resolve with answers at back.
   I have this on Forward Chess. Also on Chessable.
   https://www.chessable.com/blog/prophylaxis/

2. How to Defend in Chess, Colin Crouch, 2000
   Focus Lasker and Petrosian, about 50 annotated games.

3. Secrets of Chess Defense, M. Marin, 2003 (More advanced I think)

4. Karpov Move by Move, Sam Collins, 275pp, 2015
   Ch 1 Middlegame Themes Prophylaxis plus the games are full of 'prop' moves.

5. Petrosian Python Chess Strategy
   Selected annotated games (120)
   I saw a mention that Dvoretsky, Yusupov and Aagaard produced some good advanced Prophylaxis content.