You are trying to be helpful, but there is such a thing as too much handholding. The more helpful you are, the more likely students are to exploit this and to become a help vampire.

You do not owe your students any success. You owe them a fair chance at succeeding. But this chance is theirs to take. The thesis is solely their responsibility, especially since the purpose of a masters thesis also is to demonstrate the ability of doing independent academic work. By definition, success or failure of the thesis isn’t your fault. Some students will throw their chance away, regardless of how much you help them.

Things that might help in the future include clearer boundaries, and focusing your assistance more on methodological aspects.

• Your student currently expects responses within minutes. Matters are rarely so urgent that they need responses the same day.

  When supervising a thesis, it might be best to minimize such informal messages except for truly urgent issues, and instead have regular meetings for which the student can prepare questions. Depending on the kind of work, a cadence like one meeting every two weeks might work well. Then, you can defer any small issues that come up: “let’s discuss this at our next meeting”.

• You mentioned that you “sent many tasks”, provided material, and tried to set deadlines.

  This is fine when working with an intern or research assistant, but not for a student who is supposed to independently write a thesis. Firstly, because you’re doing their work for them. Secondly, because it’s their thesis and they should use whatever working style they believe works for them.

  Instead, it might be best to focus your assistance on methodological aspects, so that the student knows how to write a good thesis. For example, many students need an explanation about how to find useful literature, how to structure a thesis, and maybe where to find techniques that weren’t covered during lectures. A master thesis is often the longest independent project done by the student, so sharing experience with time management can also help.

  But mostly, a Socratic approach is useful, where you ask the students about their plans for the thesis. What are their goals until the next meeting? Do they feel they are on schedule, if not how can they adapt their plans? Have they considered the connection with $related_topic? How to they intend to mitigate a certain risk you are concerned about? What challenges are they currently facing, and how do they intend to solve them? You can suggest things, but it will be the student’s decision what they do with that suggestion.

  Added benefit: if you set out a plan and they fail, they will blame you, and perhaps rightly so. If they set out a plan and they fail despite knowing your concerns, that’s clearly on them.

Since the thesis in question is already close to its end, it will be difficult to switch to clearer boundaries and to more passive support. But if at least two weeks or so are left, not impossible, if you are willing to put your foot down. For example, consider an email along the following lines:

It is great that you have these questions, but this is your thesis and working through such challenges is part of it. I don’t have the time to discuss your progress right now, but I can offer a meeting $in_4_business_days. Please continue work on your thesis in the meanwhile, and we can discuss remaining issues then.

Is this impolite? Maybe, but so is bombarding you with endless requests for help.

What I find rather odd in all of this is that it’s the first thesis you supervise, and already as the main supervisor. Supervision is a skill that can be learnt, but ideally by shadowing another supervisor for one or two theses. I’m still in that “apprenticeship” phase. So I think it’s completely unsurprising that you’re running into these kinds of problems, and I also ran into similar problems with the first students that I mentored. The good news for you is that you can learn from this experience, and can be a much better supervisor for the next student. The learning experience goes both ways.