The AI Getting Things Done Coach skill for Claude
You dump everything floating around your head into Claude and get it back as a system: next actions by context, projects with their first physical step, calendar items with a date, a waiting-for list, and the things you might want someday kept separate. Every item goes through the same decision process from David Allen's Getting Things Done, vague subjects get probed until there is an action you can literally carry out, and you also get the weekly review that keeps the system accurate. Anything that takes less than two minutes does not go on a list, because then you simply do it.
the lines above come from the zip on this page · SKILL.md is 8,093 bytes
Head in, lists out
reg. G.001This is the example that appears literally in the SKILL.md, written out in full here. Eight things someone had in their head, jotted down in random order, from a quote to a trip to Japan. Watch what happens to each item: not every item becomes a task, and not a single item stays as a bare subject.
What the AI Getting Things Done Coach is
The AI Getting Things Done Coach is a free skill from our own skill library. A skill is an instruction file, SKILL.md, that gives an AI assistant a fixed way of working for a task. No software to install, no subscription, no link to your calendar or your task list: a 1,259-word text file that tells Claude how to take apart a head full of loose thoughts, which questions it asks for each item, in what order that happens and what it must never, under any circumstances, put on a list.
You download the zip at the top of this page, put it in your Claude environment and from that moment Claude processes your inbox according to this logic. What a skill actually is and why that is different from a loose prompt is explained in full in what are Claude skills NL.
The problem this skill solves is probably the reason you are here. You do not have too little time, you have too many open loops. Five things are on a scrap of paper, three in an app, two in your email and the rest is floating somewhere between your ears. Because you never have the complete picture anywhere, you stop trusting your own list, and a list you do not trust is one you eventually stop reading. The SKILL.md names that as the core of the problem: your head is not a good place to store things, and stress and procrastination almost always come from open loops floating around that are not being tracked anywhere.
This coach is written for people juggling several things at once who have lost the overview: business owners with projects alongside client work, freelancers who do their own admin, team leads with a double diary, and anyone who feels they are behind on everything. The file's description explicitly names that as a trigger, together with phrases like I keep forgetting things, my head is full and I do not know what to do first. It is part of the library of 100 free skills NL that we make available from the AI and automation service, with no account and no sales email afterwards.
There is one thing you need to know upfront. It sorts, it does not work for you. That distinction sounds trivial and in practice it changes everything. It turns a chaotic head into an organised system with concrete first steps, but you still have to take the steps yourself. And it stores nothing: you get structured text back that you put somewhere yourself, in the app you already use or in a document. Not a shortcoming, but precisely the reason the file is so small and works everywhere.
Why most task lists get stuck
Anyone who never finishes their list usually looks for the cause in discipline. It is almost always in the wording. The rules the skill sets itself read like a list of everything that goes wrong in practice, so we go through them one by one.
Your list has subjects on it, not actions. Build a website is not a task but a project. Something with the blog is not even that. Your brain cannot execute a line like that, so it skips it, and after being skipped three times it becomes part of the furniture. A hard rule applies here: an action is always a physical act you can literally carry out, and a project is anything that takes more than one step. So for every project it keeps asking until it reaches the first physical step.
Everything is dumped in one pile. Calls, emails, errands and thinking work are all mixed together, even though you can never do them at the same moment. GTD solves that with contexts: practical places or states in which an action is possible. As examples the file names @calls, @computer, @office, @home, @errands for shopping on the go, @low energy, @high energy, @with partner and @with team. If you are standing in a queue with ten minutes and a phone, you open @calls, not your whole list.
Small things cost double the work. Writing down something that takes ninety seconds, finding it again later, reading it again and then still doing it costs more time than simply doing it straight away. That is why the two-minute rule in this file is not a tip but an instruction: if the next action takes less than two minutes and fits your current context, you do it now and do not put it on a list.
There is no place for things that are not an action. An idea for two years from now, an article you want to keep, a thought that does not need to go anywhere: without a place for them, they keep floating around among your real tasks. GTD gives them their own destination, namely Someday/Maybe, Reference and Trash, and that separation is exactly why the action list stays readable afterwards.
Nobody checks the system. Every list goes stale. Projects slip, delegated things do not come back, and after two weeks your overview no longer matches reality. That is why the SKILL.md requires the skill to include a weekly review with every plan, and that review is concrete: go through the calendar two weeks back and two weeks ahead, go through every list, put the project list next to the next-actions list so every project has at least one next action, clean up the waiting-for list and do a fresh mind sweep for whatever has come in since last week.
In diff form, with lines from the skill's example: what goes out and what comes back in its place.
Six questions, fixed order
reg. G.002Step 2 of GTD is called Clarify, and that is where the work happens. The SKILL.md sets out exactly six questions, in this order, and every item goes through them without exception. The questions do the thinking, so you do not have to redo it every time.
What is actually in the SKILL.md
A skill is only as good as its instructions, so we simply describe them here. At the top is a frontmatter with the name, the version and the licence, together with the description that determines when Claude picks up the skill. That description is remarkably broad.
Not just the tidy terms like Getting Things Done, GTD, mind sweep, weekly review, next action and inbox zero, but also the way people actually say it: I keep forgetting things, my head is full, too many open tasks, I do not know what to do first, empty my head. There is even an explicit instruction to trigger for business owners running several projects at once who feel they are behind on everything.
After that comes a short paragraph on sources and theory, and immediately after that the method, which must be followed in order without exception. Step 1 is Capture: the user writes down everything on their mind, without filtering and without sorting, work and personal mixed together. If the user already supplies a list, the file states literally that the list is taken over without reinterpretation, because only that way are you working with a real inbox rather than a summary of one.
Step 2 is Clarify, with the six questions from FIG.02 above. Step 3 is Organise: every processed item goes to one of the seven places shown in FIG.03 below. That step also carries the instruction on contexts, with nine examples and the explicit order to use whatever fits the user's life and never to invent labels.
Step 4 is Reflect, and in practice this is the step that most often falls by the wayside. That is why it must tell the user that the system only works with a weekly review, and must include a concrete weekly-review prompt with every plan. Step 5 is Engage, with a simple decision logic of four criteria, in this order: what can I do in this place and with this tool, how much time do I have, how much energy do I have, and what has the most impact on my projects and roles. Priority is deliberately placed last, because an important task you cannot do here and now is not the right task right now.
The file then fixes the output in eleven blocks with set names, from What I see in your inbox to First physical action (NOW), so every answer looks the same and you can lay two sessions side by side. It closes with the worked example from FIG.01, five lines under the heading Important rules, and three sources. If you want to learn to build a file like this yourself, writing a SKILL.md NL explains the structure step by step.
Seven places, there are no more
reg. G.003Step 3 of GTD is Organise, and the SKILL.md is strict about it: anything that is a commitment belongs on one of these seven places, and nowhere else. Hidden lists are forbidden, because you do not use a system you do not fully trust. Under each place is what went there in the example from FIG.01.
The theory the skill rests on
The SKILL.md names three sources explicitly, and that is deliberate: so you can look them up and decide for yourself whether you agree. Together they explain why the decision process looks the way it does.
The system comes from David Allen. Getting Things Done was published in 2001 and got a revised edition in 2015. The starting point of that book is that your head is not a good place to store things. It is excellent at coming up with ideas and judging them and downright bad at storage, and yet most people mainly use it as storage. The five steps are how Allen moves that storage outside your head, with the emphasis on the question of the next physical action.
The detail rests on the GTD Methodology Guide. The second source in the file is the methodology guide from the David Allen Company. That is where the practical details come from that make the difference between a nice idea and a working system: the contexts, the split between projects and actions, and the design of the weekly review.
The explanation comes from Masicampo and Baumeister. The file cites their 2011 research, published in the Journal of Personality and Social Psychology, on the Zeigarnik effect: unfinished tasks keep demanding attention. The title of that paper is telling, because its core finding is that making a plan lowers the mental load of an unfinished task. Not finishing it, then, but writing it down and deciding. That is exactly what steps 1 and 2 of GTD do, and it is why a mind sweep brings relief while not a single task has actually been finished yet.
Even without installing the skill you can do something with this today: write everything down, ask for each item whether it requires action, formulate the first physical step, and do everything that takes less than two minutes straight away. What the skill adds is that it takes over that work and keeps the order, even on the days you do not feel like it yourself. Those are exactly the days that matter.
What the skill refuses
reg. G.004Under the heading Important rules are five hard rules, and a few more are spread through the method. Together they determine what the skill does not do, and that matters at least as much as what it does do. In conversation those rules play out like this: every row is a request you might make, with the response that follows from the file.
Installing in Claude Code, Claude.ai or Codex
The zip contains a folder with the SKILL.md inside it. Installing is a matter of putting the file in the right place, and that place differs per environment. SKILL.md has been an open standard since December 2025, so the same skill also works in Codex, Cursor and Gemini CLI. So you are not downloading a Claude file but a working instruction that any modern AI assistant can read.
- Unzip it into
~/.claude/skills/(or.claude/skills/in your project). - Claude then recognises the skill automatically as soon as you start talking about tasks, overview or a full head.
- You can also call it directly, with
/ai-getting-things-done-coach.
- Go to Customize and then Skills.
- Upload the zip there as a skill.
- Or paste the contents of SKILL.md into the project instructions of a Project.
- Open
AGENTS.mdin your repo. - Paste the contents of SKILL.md into it, or put SKILL.md next to it as a separate file and refer to it from
AGENTS.md. - Codex reads that along with every session.
If the installation gets stuck anywhere, or you want the steps with screenshots, the full explanation is in installing Claude skills NL. After that, using it is simple: type everything on your mind into a message, without sorting it, and let the skill do its work. The messier the dump, the more it delivers. If you are new to this kind of work and want the wider context first, you will find the explanation of working with AI in the knowledge base.
When to use it, and when not to
This coach is at its strongest the moment things have genuinely become too much. A Monday morning when you do not know where to start, the day after a holiday, the end of a quarter, or the moment you notice you have put off the same thing three times without knowing why. It is also strong as a recurring ritual: repeating the mind sweep once a week and running the weekly-review prompt from the previous session again, so the system stays accurate instead of slowly going stale.
There are also situations where you are better off leaving it alone. For choosing between two roughly equal projects GTD does not help, because the method organises what is there and does not weigh up what it is worth. If you want to set urgency against importance, the Eisenhower Matrix Coach skill is the tool for that, and the background on that model is on the page about the Eisenhower matrix NL. If the problem is not in your list but in the goals behind it, start with the SMART Goals Coach skill, because a list cannot fix the fact that you are working on the wrong things.
A second limit is behaviour. GTD tells you what the next step is, not how to get yourself to take it. If you keep putting off actions that are clear and small, the obstacle lies in habits, not structure, and the Atomic Habits Coach skill is a more logical starting point. If you do manage to start but keep getting interrupted, it is about concentration, and the Flow State Coach skill takes you further.
And the most important honest limit: the skill is not a task-list app. It stores nothing between sessions, it sends no reminders and it does not look at your calendar. What it delivers is structured text that you put somewhere yourself. That is exactly why it works everywhere and costs nothing, but it also means the upkeep stays with you. If you want the upkeep itself taken off your hands, the conversation continues in the next section.
Run it yourself or have it run
reg. G.005This skill is the free do-it-yourself version of work we also deliver as a service. It stays complete and with no catches, but be clear about what a skill is: it teaches your AI how to do something, while every new session starts empty. It is not the engine and not the memory. You dump, you supply the context again every time, you check the result. Anyone who wants that differently has two follow-up steps: hand over the engine, or sort out the memory.
where you are now The skill: you are the engine You run the AI Getting Things Done Coach yourself in Claude, Codex or Cursor. Costs nothing, works today, and you keep it entirely in your own hands: no trial period, no locked-off parts. The limit is your own time: it only happens when you do a mind sweep and put the result somewhere.
only if you come back The AI employee: usually not the next step here None of our roles fit here, and we would rather say that than sell it to you. The three roles we set up ready-made are the Quote Employee (sorting incoming requests and preparing draft quotes), the Sales Employee (prospect research and outreach drafts) and the Reporting Employee (summaries and weekly and monthly reports from your own data). What this skill delivers, an empty inbox and next actions by context, fits none of those three. What is possible is a custom-built role via Mansotti, the company of which TheSEO is the trading name, but only if it concerns a team's inbox and not just yours. If this work stays a single session a year for you, skip step 2: you run your own system yourself; what does help is an AI that remembers your lists between sessions, and that is step 3. What an AI employee does do is on that page.
everything from one source Jarvis: all your AIs work from the same company knowledge The skill teaches the AI, the brain is where the memory lives. Want all your AIs to work from the same company knowledge: that is Jarvis, the organisation's brain. It connects ChatGPT, Claude, Codex and your people to the same projects, core knowledge and decisions, so your next AI session does not start from zero. Jarvis selects context, delegates work and stores only what has been approved: client knowledge stays isolated and every step leaves a checkable trail. This skill specifically gets better because of it, since you no longer have to supply your projects, your contexts and your waiting-for list again every session. What that delivers in practice, from the plans to your first week, you can read at Jarvis itself.
What Jarvis actually delivers
reg. G.006Step 3 deserves more than one paragraph, because this is the difference between a clever chat and a system you can build on. Jarvis is the organisation's brain: it remembers what your AIs need to know, divides up the work and keeps track of what happened. With GTD that counts double, because a system that has to be re-explained every week is, by definition, not a system. You notice it first at the start of a new session.
We have been running on this system ourselves for months. Every agent session, every task and every decision is logged in it and can be read back. A new session therefore does not start blank: it first retrieves the recorded decisions, the running projects and the latest changes, and carries on where the last one stopped. So we are not describing a promise but the way of working we operate in ourselves, every day.
See the four plans at jarvis/pricing NL. Through the waitlist NL you only pass on your preferred plan, without obligation. That does not yet create an account, an order or a payment obligation. We discuss bespoke business work first.
The skills around it
reg. G.007GTD organises what is there. It does not choose for you, it does not motivate you and it does not decide where you want to go. These skills from the same library each take on a different part of that chain.
Choosing what comes first
the weighingYour list is complete, but it is still too long. These two help with the order.
Eisenhower Matrix CoachWeighs urgency against importance, exactly the weighing GTD deliberately does not do.SKILL SMART Goals CoachTurns a direction into a goal you can check, so your project list is actually about something.SKILLActually doing it
the behaviourA perfect list does not fix procrastination. These two are about execution.
Atomic Habits CoachFor actions that are small and clear and still do not get done: then the problem is habit.SKILL Flow State CoachFor when starting works but continuing does not, because everything interrupts you.SKILLGetting better at the work itself
the craftOnce the list is right and you work through it, it becomes about the quality of the work.
Deliberate Practice CoachSets up focused practice for what GTD only clears space for.SKILL Minutes WriterCaptures the action points from a meeting, so they do not end up back in your head after all.SKILLLooking further
the contextWhere this skill comes from and what else there is.
The whole skill libraryAll 100 free skills in a row, sorted by subject.HUB NL AI and automationThe service behind it: from separate skills to working automation in your business.SRV AI trainingIf your team wants to learn to set up this kind of work themselves.SRV NL Company Brain TemplateFree context file, so your AI knows your business before you ask it anything.DOC NLFrequently asked questions
What does the AI Getting Things Done Coach skill cost?
Nothing. The skill is free, the file itself states the MIT licence, and you do not need to create an account or leave an email address. You download a 3.8 KB zip containing a folder and one file, a 1,259-word SKILL.md, and that is the complete skill. There is no paid version and no sales email follows.
Does this skill also work in Codex, Cursor or Gemini CLI?
Yes. SKILL.md has been an open standard since December 2025, so the same file also works in Codex, Cursor, Gemini CLI and other tools that follow the standard. In Codex you unzip it into .agents/skills/ in your project, or into ~/.agents/skills/ for all your projects; Codex has supported SKILL.md directly since the open standard of December 2025. Putting the contents of SKILL.md into your AGENTS.md still works too. The instructions are plain readable text, so any assistant that accepts instruction files can work with it.
Do I need to read David Allen's book first?
No. The SKILL.md contains the five steps, the six decision questions, the seven places and the weekly review in the order you need them. The book is listed as a source, together with the GTD Methodology Guide from the David Allen Company, so you can always go back to the original explanation if you want to go further.
What is the difference between a project and a next action?
A project is anything that takes more than one step. A next action is a physical act you can literally carry out. Build a website is therefore a project, and the action for it is, for example, sketch a homepage wireframe at your computer. The skill refuses to put a project on the list as an action and asks for the first next action for every project.
Which contexts does the skill use?
Contexts are practical places or states in which you can carry out an action. The file gives calls, computer, office, home, errands for shopping on the go, low energy, high energy, with partner and with team as examples. The instruction is explicit: use what fits the user's life and do not invent labels.
Does the skill replace my task list or my calendar?
No. The skill is a text file that does the sorting, not an app that stores anything. You get the outcome back as structured text: next actions by context, projects with a first action, calendar items, waiting for, someday maybe, reference and the two-minute quick wins. Where you keep those lists afterwards remains your choice.
What does the skill do with things I might want someday?
Those go on the Someday/Maybe list, the place for ideas with no commitment. In the example in the file that happens with a trip to Japan and with a tool the user wants to look at. So they do not disappear, but they also do not sit among the actions you have to do something with today.
A clear head that stays clear
A clear head is a good start, but GTD stands or falls on the weekly review, and that is where most systems run aground. Put that review in your calendar before you start on the rest.
With a stand-alone skill you run into that: every session starts empty, so you supply your lists, your contexts and your open actions again every time. That is exactly the problem Jarvis solves, from €9 per month including VAT. If it is not about your inbox but about a whole team's, having the work prepared for you becomes a real option, and that is the AI employee.