“Be my life coach” is too vague. It usually produces agreeable advice, a long list, and no behavioral commitment.
You do not need to read all 25 prompts. Match the problem you have right now, then use the setup prompt and one numbered prompt.
Pick your prompt in 15 seconds
- Your goal is vague: Start with prompt 1.
- Your week keeps drifting: Use prompt 6 for a review or prompt 8 for repeated carryover.
- A habit will not stick: Use prompt 11 to shrink it or prompt 15 to recover after a miss.
- You are stuck on a decision: Use prompt 16 for a reversible test or prompt 20 to challenge your preferred option.
- Your self-talk is making the problem worse: Use prompt 21 or 22.
- You need follow-through: Use prompt 24 to make the commitment, then prompt 25 at the next check-in.
First session? Copy the setup below, then paste the numbered prompt that matches today. Stop when you have one action with a time and a deadline.
Copy this first: the 60-second coaching setup
Copy this into a new chat and replace the brackets.
Act as a direct, practical life coach. Your job is to improve my decisions and follow-through, not to agree with me. Use the context I provide as evidence. Identify vague language, avoidance, conflicting priorities, and assumptions that are not supported by facts. Ask one question at a time when you need me to think. Do not give me a large list of options. Recommend the most useful next move and explain the reasoning briefly. End each session with exactly one action, one trigger or calendar time, one deadline, and one way to verify completion. My current goal: [goal] Why it matters: [reason] My real constraints: [time, energy, money, health, or responsibilities] What I have done in the last seven days: [behavior, not intentions] What I keep avoiding: [truthful answer] Begin by stating the main pattern you see in two sentences. Then ask your first question.
That prompt establishes the operating rules. The prompts below give the session a specific mission.
ChatGPT product screen captured July 2026 with a TaskCoach test prompt.
How to use these prompts
- Pick one prompt, not five.
- Replace every bracket with specific facts.
- Answer follow-up questions with behavior from the last seven days.
- Reject any action you cannot observe or schedule.
- Put the final commitment somewhere you will see it again.
The standard is simple: if the session does not change what you do next, it was an interesting conversation, not coaching.
Goal clarity prompts
1. Turn a vague goal into a measurable target
My goal is “[goal].” Convert it into one measurable outcome with a deadline. List the assumptions hidden inside my wording, choose the most realistic definition of success, and give me the first action that can be completed in 30 minutes or less.
2. Find the real goal behind the stated goal
I say I want “[stated goal].” Ask me one question at a time to identify what I actually want this goal to change in my life. Stop after five questions. Then state the deeper goal, the evidence for your conclusion, and one way my current plan may be solving the wrong problem.
3. Choose one goal from competing priorities
I am choosing between [goal A], [goal B], and [goal C]. Score each option from 1 to 10 on impact, urgency, reversibility, effort, and alignment with my stated values. Show the scores in a table, choose one priority for the next 30 days, and state what I must deliberately pause.
4. Convert an outcome into controllable inputs
The outcome I want is [outcome]. Separate what I control, influence, and cannot control. Build a weekly input target using only controllable behaviors. End with one action I can complete today.
5. Expose the hidden cost of the goal
I want to achieve [goal] by [date]. Calculate the likely weekly time, money, energy, and opportunity cost based on the details below: [details]. Identify the constraint most likely to break the plan. Redesign the goal so the cost is explicit and sustainable.
Weekly review prompts
6. Run a no-excuses weekly review
Here is what I planned last week: [plan]. Here is what actually happened: [results]. Separate facts from explanations. Identify the single pattern that explains the largest gap, then prescribe one change for next week. Do not redesign my entire life.
7. Find the highest-leverage win
My completed actions this week were: [list]. Identify which action produced the most useful result per unit of time or energy. Explain why it worked and design one repeatable version for next week.
8. Diagnose repeated carryover
I postponed these items more than once: [list]. For each item, classify the likely cause as unclear next action, low value, missing resource, unrealistic size, fear, or scheduling conflict. Recommend delete, delegate, break down, or schedule. Choose one classification when the evidence is strongest.
9. Compare my calendar with my claimed priorities
My stated priorities are [priorities]. My actual time use was [summary or calendar]. Identify the largest mismatch. Do not judge the intention. Judge the allocation. Give me one calendar change that would make next week more honest.
10. Build next week from evidence
Use this week’s energy, interruptions, wins, and misses to design next week: [evidence]. Keep the plan within my real capacity. Choose one primary outcome, three supporting actions, and one rule for what I will not add.
Habit design prompts
11. Make a habit too small to avoid
I want to build the habit [habit]. My current consistency is [frequency]. Reduce it to a version that takes under two minutes and still counts as a valid repetition. Attach it to an existing daily event and define exactly when I record completion.
12. Create an implementation intention
Turn this intention into a precise if-then plan: [intention]. Use the form “If [specific cue or obstacle], then I will [specific behavior] at [place or time].” Add one recovery plan for the most likely disruption.
13. Remove friction from a good habit
The habit is [habit]. The moment I usually fail is [moment]. Map every step from cue to completion. Identify the step with the highest friction and redesign the environment so that step becomes easier before motivation is required.
14. Add friction to a bad habit
The behavior I want to reduce is [behavior]. Its usual cue, location, device, and reward are [details]. Design three layers of friction: one immediate barrier, one environmental change, and one accountability consequence. Choose measures I can verify.
15. Recover after a missed habit
I missed [habit] on [date] because [what happened]. Do not motivate me. Diagnose whether the failure came from the cue, ability, environment, plan size, or competing demand. Preserve the habit by designing the smallest valid recovery action for the next scheduled opportunity.
Decision-making prompts
16. Make a reversible decision quickly
I am stuck on [decision]. First classify it as reversible or hard to reverse. If it is reversible, give me a 48-hour experiment with a clear success measure. If it is hard to reverse, identify the missing information that would most change the choice.
17. Run a pre-mortem
Assume my plan to [plan] failed six months from now. List the five most plausible causes, ranked by likelihood and impact. For the top two, add one prevention measure and one early warning signal.
18. Separate fear from evidence
I am worried that [fear]. Create two columns: evidence that supports the fear and evidence that weakens it. Mark assumptions as assumptions. Then recommend the smallest safe test that would produce new evidence.
19. Use regret minimization
I am choosing between [options]. Compare likely regret after 10 days, 10 months, and 10 years. Separate regret caused by a bad outcome from regret caused by avoiding a meaningful attempt. Recommend one option and state the strongest argument against your recommendation.
20. Challenge my preferred option
I currently prefer [option] because [reasons]. Argue the strongest case that I am wrong. Identify confirmation bias, sunk cost, social pressure, or convenience if present. Then state whether the original choice survives the challenge.
Self-talk and accountability prompts
21. Rewrite a harsh thought without lying
The thought is: “[thought].” Identify the cognitive distortion, if any. Rewrite it as a statement that is accurate, specific, and useful without forced positivity. End with one behavior that would create better evidence.
22. Convert shame into a repair plan
I feel ashamed about [event]. Separate what happened, what I am making it mean about me, who was affected, and what can still be repaired. Give me one repair action and one boundary that prevents repetition.
23. Call out my excuse
I keep saying “[excuse]” about [goal]. Treat the statement as a hypothesis. Identify what part is fact, what part is avoidance, and what action would still be possible if the constraint remains true. Be direct.
24. Create an accountability contract
I commit to [action] by [deadline]. Write a short accountability contract with a clear definition of done, proof of completion, a check-in time, and a reasonable consequence if I do not follow through. Keep it proportionate and safe.
25. Run the next-session check-in
At our last session I committed to [action] by [deadline]. The result was [completed, partly completed, or not completed], and the evidence is [evidence]. Evaluate the plan, not my worth. Decide whether to repeat, reduce, replace, or remove the action. Then issue the next commitment.
Why these prompts work better
The prompts share four mechanisms.
They use evidence. “I want to exercise more” gives the model almost nothing. “I planned three sessions, completed one, and skipped both evening sessions after late meetings” exposes a pattern.
They constrain the output. A request for “ideas” invites a list. A request for one diagnosis and one action forces prioritization.
They separate decisions from identity. A missed action is data about the plan, cue, capacity, or environment. It is not proof that the person is lazy.
They create implementation intentions. Research by Peter Gollwitzer showed the value of linking a future situation to a specific response. “If the meeting ends after 18:00, then I will do the 10-minute version at home” is more executable than “I will try harder.”
What no prompt can guarantee
Prompt quality is not system architecture.
- Context: A chat can only use the information, memory, and connected tools available to it. Important facts may be missing or stale.
- Observation: The model may not know what happened in your calendar, habit tracker, or real environment unless you provide or connect that data.
- Follow-up: Reminders and proactive check-ins depend on product features and how you configure them. A written commitment does not enforce itself.
- Accuracy: A confident answer can still be wrong. Verify medical, legal, financial, and other high-stakes advice with qualified sources.
Use ChatGPT for thinking, reframing, planning, and structured review. Put the final action into a system that will surface it at the right time.
Where TaskCoach fits
TaskCoach removes some of the manual setup by connecting coaching with goals, habits, notes, tasks, and a schedule. Its nine coach personas already provide distinct coaching stances, and approved commitments can move into the same planning system used for follow-through. That is useful when maintaining the context block becomes more work than the coaching session.
In this real session, Sky checks Mara's recent completion data, identifies that messaging practice is ahead of real-time conversation practice, and recommends three short social reps each week.
Sky uses actual habit history to answer "How am I doing?" and recommend the next routine.
The recommendation does not disappear into chat. Sky prepares a specific Sunday habit for review, and Mara approves it before it enters her system.
The Brain Agent turns its recommendation into an applied habit with a visible approval step.
Start with the prompts above. If they solve the problem, keep using them. If the repeated failure is memory, scheduling, or follow-up, compare the broader systems in our guide to the best AI life coach apps.
Your order
Copy the master prompt. Pick one of the 25 prompts. Complete the action before collecting another piece of advice.