todoerTurn a vague goal into a useful task tree, with a worked example and a review-first AI prompt.
Start with a finish line
A project describes a result that requires several actions. A task describes work you can begin. “Website” is a subject; “Publish a portfolio with three case studies” is a finish line. Write the outcome before making a checklist so that the checklist has a reason to end.
Define what is outside the scope as well. Your portfolio project might exclude a blog, a new logo and a full rewrite of every case study. Without a boundary, decomposition becomes a way to discover endless additional work instead of finishing something.
Break the outcome into deliverables
Use the first level for meaningful pieces of the result. For a portfolio, those might be case studies, homepage and publishing. Use the next level for actions: select screenshots, draft a case-study summary, check links and publish. Keep each child attached to the deliverable it helps complete.
Do not split every action into microscopic steps. “Open the editor” rarely needs a task. Add another level when it reveals a dependency, a separate decision or work that requires its own context. The tree should help you begin, not describe every movement of your hands.
Write tasks that can survive a week away
Use a verb and a concrete object. “Review” leaves a future version of you guessing. “Check the homepage on a phone and record layout issues” gives you an action and a completion condition. Add links, acceptance notes or files beside the task when they reduce the effort of restarting.
A good test is to close the project and return later. If the title alone makes sense but you still need to search five places for context, improve the description. If you cannot tell whether the task is complete, make the finish condition more specific.
Worked example: a small portfolio launch
Outcome: publish a three-case-study portfolio. Deliverable one: choose three projects and collect approved screenshots. Deliverable two: draft the homepage headline and summaries. Deliverable three: check mobile layouts, fix broken links and publish. This is a useful initial tree because it exposes the biggest work without pretending every detail is known.
The first next action is not “Build portfolio.” It is “Choose three projects.” After that decision, new actions become obvious. Avoid generating a hundred speculative tasks before the first decision; that creates a large plan whose assumptions have not been tested.
Ask AI for a draft, not a verdict
Provide the outcome, constraints and things you already know. Ask the model to identify missing decisions and label assumptions. A generic “break this down” prompt often produces attractive but unnecessary work. You need a plan for your project, not an average plan for a similar-sounding project.
Try: “Draft a task tree for publishing a three-case-study portfolio. I have existing screenshots and two evenings. Exclude branding and a blog. Keep the first draft under twelve tasks. Mark assumptions and propose the first action. Do not write to my task manager.” Edit the result before saving it.
Know what completion means in your tool
Nested tasks can behave differently across apps. In Todoer, an unfinished descendant blocks ordinary parent completion; explicit subtree completion is a separate action. Reopening a child reopens its ancestors. These rules help distinguish finishing a parent from deliberately closing an entire branch.
Before bulk-completing or moving a task tree, inspect its children. Dates, permissions and project context may matter more than the top-level title suggests. Start with a small project to learn the behavior, then use the same structure when the stakes are higher.
Keep the tree alive without gardening it constantly
During a weekly review, remove obsolete branches and split tasks that remain too vague to start. Leave useful structure alone. If you repeatedly reorganize a project without doing it, identify the decision you are avoiding or reduce the scope to one deliverable.
The practical measure is whether the structure lowers the effort of the next work session. A shallow tree with clear actions is more useful than a perfectly categorized tree nobody returns to. Use the weekly review guide to keep active projects realistic.
Give one next step a home.
Start with an inbox and a real task. Add structure when it helps.
Create your account ↗