Claude Prompt Library

Claude Ran My Evenings — 50 Prompts to Plan Tomorrow in 90 Seconds

50 copy-paste prompts

50 Claude prompts for tomorrow's plan, today's review, evening journal, emails to send tonight, and one decision reframed in two sentences — before bed.

In short: This page contains 50 copy-paste ready prompts, organized into 5 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

Louis Corneloup
By Louis Corneloup · Founder, Techpresso
Last updated ·Hand-curated & tested by the AI Academy team

Tomorrow's Plan

10 prompts

Tomorrow at a Glance

1/50

✨ What it does

Claude writes a one-page brief for tomorrow with one top priority, a focus block, and three emails to send tonight. You paste the calendar, mark energy as [low / normal / high], then follow the brief or edit the priority.

<task>Plan tomorrow in a single-page brief.</task> <inputs> - Tomorrow's calendar: [paste] - Tomorrow's open commitments: [list] - One thing I'm sleeping on: [describe] - My energy expectations tomorrow: [low / normal / high] </inputs> <output> A "Tomorrow at a Glance" doc with: 1. Top priority — the one thing that, if done, makes tomorrow a win. 2. Focus block — protected 2hr window with what to do in it. 3. Three emails to send tonight (with subject lines) so I'm not scrambling at 9am. 4. Decision reframed in 2 sentences so I can actually sleep. 5. One thing I will say no to tomorrow. 6. End-of-day reflection question to answer tomorrow night. </output>

💡

Pro tip: The "3 emails to send tonight" pattern is the highest-leverage evening practice — 5 minutes tonight saves 30 minutes of scramble tomorrow morning.

Reorder Tomorrow by Impact

2/50

✨ What it does

Claude reorders tomorrow's tasks by impact, putting high-impact work in your sharp hours instead of by urgency. You paste the task list and this week's main outcome, then run the day in that order.

<task>Reorder tomorrow's tasks by impact, not urgency.</task> <tasks>[list with rough time estimate each]</tasks> <context> - This week's main outcome: [describe] - My energy curve: [when I'm sharpest] </context> <output> 1. Reordered list — high-impact tasks at my sharp hours, low-impact in low-energy windows. 2. Tasks that look urgent but aren't — flag for delegation or deletion. 3. Tasks I keep putting off — name the resistance. 4. Tomorrow's first 90 minutes — exactly what to do, no decisions needed. </output>

💡

Pro tip: Most people start the day with email and finish with strategic work — exactly backwards. The "first 90 minutes" prompt fixes that by removing morning decision fatigue.

Calendar Triage

3/50

✨ What it does

Claude triages tomorrow's meetings and marks which to cancel, shorten, or send someone else to, plus what prep you need. You paste the meeting list, then send the cancels and keep the rest.

<task>Triage tomorrow's calendar.</task> <calendar>[paste meeting list]</calendar> <output> For each meeting: 1. Purpose: status / decision / generative / social. 2. Cancel candidate: yes / no with reasoning. 3. Shorten candidate: 60→30 or 30→15. 4. Delegate candidate: someone else could attend. 5. Prep needed: what I should review beforehand. Then: my "version of tomorrow" if I executed all the cancellations and shortenings. </output>

💡

Pro tip: Most calendars have 20-30% slack hidden in optional or shortenable meetings. Run this every Sunday night and Wednesday night — reclaim 5+ hours per week.

Meeting Prep Brief

4/50

✨ What it does

Claude writes a prep brief for [specific meeting tomorrow] with the one outcome you want, questions to ask, and a prepared no. You fill in attendees and stakes, then walk in with that brief.

<task>Brief me for [specific meeting tomorrow].</task> <meeting> - Title: [name] - Attendees: [list] - Context: [what I know about why it's happening] - Stakes: [what's at risk or in play] </meeting> <output> 1. The single most important outcome I want. 2. The single most important thing I want to learn. 3. 2 questions I should ask. 4. 2 things I should NOT say. 5. The version of "no" I might need ready. 6. Energy level needed (so I can pace the rest of my day). </output>

💡

Pro tip: The "prepared no" is the highest-value meeting prep. Most yeses you regret get given because you didn't pre-decide where the line was — Claude pre-decides for you.

Tomorrow's One Big Thing

5/50

✨ What it does

Claude picks the single most important task for tomorrow, explains why not the others, and notes where to resume if you get interrupted. You list what is on your plate and the hours you have, then protect that one block.

<task>Help me pick the One Big Thing for tomorrow.</task> <context> - This week's goal: [describe] - Stuff on my plate: [list of options] - Time I'll actually have: [hours] </context> <output> 1. The One Big Thing — picked, with rationale. 2. Why not the others (briefly). 3. Time block where it goes. 4. The "if interrupted, here's where I'd resume" note. 5. Defining done — what completed looks like. </output>

💡

Pro tip: The One Big Thing discipline beats long task lists. Most productive days have one breakthrough output, not ten medium ones — Claude forces the picking.

Pre-Mortem on Tomorrow

6/50

✨ What it does

Claude imagines tomorrow going wrong, names the most likely derailer, and lists moves you can make tonight. You list priorities and known risks, then set those guards before you sleep.

<task>Run a pre-mortem on tomorrow.</task> <plan> - Top 3 priorities: [list] - Meetings: [list] - Known risks / blockers: [list] </plan> <output> 1. Imagine tomorrow ends and you got nothing done. What happened? 2. The single most likely derailer (interrupting meeting, urgent ask, energy crash). 3. Pre-emptive moves I can make tonight to prevent it. 4. Plan B: what's the minimum-viable version of tomorrow if things slide? 5. Permission slip: what I'm allowed to skip if needed. </output>

💡

Pro tip: Pre-mortems beat optimism. Identifying the derailer before it happens lets you set up defenses tonight — usually a single declined meeting or one phone-off block.

Block & Tackle Plan

7/50

✨ What it does

Claude splits tomorrow into block hours for deep focus, tackle hours for meetings and admin, and buffer time between them. You paste your calendar and priorities, then honor the blocks.

<task>Build tomorrow's block-and-tackle plan.</task> <inputs> - Calendar: [paste] - Priorities: [list] - Energy notes: [when I'm sharpest] </inputs> <output> 1. Block hours: deep-focus windows for priority work. 2. Tackle hours: meetings, slack responses, admin. 3. Buffer hours: 30-min cushions between block and tackle. 4. Email windows: 2 specific times, max. 5. Hard stop: when I close work. </output>

💡

Pro tip: Block-and-tackle separation produces 2-3x the deep work of mixed-mode days. The hard stop matters too — the brain works better with a known end.

Weekly Re-Plan Sunday Night

8/50

✨ What it does

Claude replans the week from where you are now, keeping what still matters and dropping what has been overtaken. You paste original priorities and what shifted, then use the new list on Monday.

<task>Re-plan the week from where we are now.</task> <state> - Week started with: [original priorities] - What's shifted: [describe] - Calendar already booked: [paste] - New things on my plate: [list] </state> <output> 1. What still matters from the original plan. 2. What's been overtaken — drop it. 3. New priorities — be ruthless, max 3. 4. The week's central question (what I'm trying to answer or move). 5. Re-planned calendar with focus blocks. 6. One thing I'll celebrate at week's end. </output>

💡

Pro tip: Sunday-night replan beats Monday-morning rush. 15 minutes Sunday saves 90 minutes of "wait, what was I doing?" Monday.

Morning Setup From Tonight

9/50

✨ What it does

Claude writes a 90-minute morning script so you can follow steps at wake time instead of deciding at 7am. You name tomorrow's first commitment and wake time, then run the script as written.

<task>Set up tomorrow morning so I don't have to think.</task> <context> - Tomorrow's first commitment: [time + topic] - Wake time: [estimate] - Anything I want to write / make / decide first thing: [list] </context> <output> A 90-minute morning routine, written so I can follow without deciding: 1. Wake → first 15 minutes (what to read / not read). 2. Next 30 minutes: deep work — exact task and starting point. 3. Next 30 minutes: top priority continued or shift. 4. Last 15: light prep for the day's first commitment. 5. What I'm NOT doing in the first 90 minutes (email, slack, news). </output>

💡

Pro tip: Decision fatigue is highest first thing — pre-scripting the morning is the single biggest productivity unlock. Most people can't resist email if they have to decide; they can resist if "no email until 9am" is pre-written.

Travel Day Plan

10/50

✨ What it does

Claude plans a travel day with work to finish before you leave, realistic work in transit, and a simple landing block after. You paste [flight / drive / train + times] and how you will feel, then follow those three blocks.

<task>Plan tomorrow's travel day so I get value, not just transit.</task> <context> - Travel: [flight / drive / train + times] - Total transit time: [hours] - Work I could do en route: [list] - Energy expectations: [be honest about how I'll feel] </context> <output> 1. Pre-travel: what to finish before leaving. 2. In-transit: realistic work (not aspirational). 3. Light-mode tasks: reading, listening, low-cognitive admin. 4. What I'm NOT doing in transit (e.g., no important emails on phone). 5. Buffer for arrival: settle / eat / shower before next commitment. 6. Tomorrow's reset evening if travel ends late. </output>

💡

Pro tip: Travel days are 50% as productive as we think. Plan for that — schedule reading, podcasts, and low-cognitive work, not deep strategic work that won't actually happen.

XML tags are just the start. Learn the full Claude workflow.

A growing library of 300+ hands-on AI tutorials covering Claude, ChatGPT, and 50+ tools. New tutorials added every week.

Start 7-Day Free Trial

Today's Review

10 prompts

End-of-Day Review

11/50

✨ What it does

Claude runs a three-minute end-of-day review with three wins, the biggest plan-versus-actual gap, and a handoff for tomorrow. You list what you planned and what you did, then close the day with that note.

<task>Run a 3-minute end-of-day review.</task> <inputs> - What I planned for today: [list] - What I actually did: [list] - Energy now: [low / med / high] </inputs> <output> 1. Wins (no matter how small): list 3. 2. The biggest gap between planned and actual — what caused it? 3. One thing I'd do differently if I rewound today. 4. One thing to carry into tomorrow. 5. One thing to drop from tomorrow (an obligation I don't need to carry). </output>

💡

Pro tip: Three-minute reviews beat 30-minute reflections for consistency. Most people skip long evening reviews — short ones become habit.

What Actually Worked Today

12/50

✨ What it does

Claude names the one or two things that produced real progress today, why they worked, and the task that only felt productive. You describe the day in bullets, then repeat the conditions that worked.

<task>Identify what actually worked today.</task> <today>[describe in 4-6 bullets what you did]</today> <output> 1. The 1-2 things that produced real progress. 2. Why they worked — environment, mindset, sequencing, energy. 3. How to replicate the conditions tomorrow. 4. The thing that felt productive but wasn't (looked busy, moved nothing). 5. One small structural change for tomorrow based on what worked. </output>

💡

Pro tip: The "felt productive but wasn't" prompt is uncomfortable and high-value. It surfaces the meeting marathon or inbox-flow that looked like work and produced nothing.

Energy Audit

13/50

✨ What it does

Claude maps your energy peaks and crashes across the day and suggests a tomorrow plan that uses the peaks. You fill each window's time and felt energy, then schedule tomorrow from that map.

<task>Audit my energy across today.</task> <context> - Wake: [time + felt energy] - Mid-morning: [time + felt energy] - Post-lunch: [time + felt energy] - Late afternoon: [time + felt energy] - Evening: [time + felt energy] - Activities at each window: [list] </context> <output> 1. Energy peaks — what was I doing, what should I do during those windows? 2. Energy crashes — what triggered, what'd prevent. 3. Mismatch flags: high-energy hours spent on low-value work, or vice versa. 4. Tomorrow's energy plan: align hardest work to my peak windows. 5. One ritual that'd boost the weakest window (lunch, walk, nap, snack). </output>

💡

Pro tip: Energy is a more useful planning variable than time. Most people have 2-3 hours of peak energy daily — losing those to email is the biggest avoidable productivity drain.

What I Almost Forgot

14/50

✨ What it does

Claude captures follow-ups you owe, decisions to write down, and weak signals you noticed but did not act on. You paste today's conversations and promises, then put those follow-ups on the list.

<task>Capture what I almost forgot from today.</task> <inputs> - Conversations today: [brief notes] - Decisions made today: [list] - Promises I made: [if any] - Things I noticed but didn't act on: [list] </inputs> <output> 1. Follow-ups I owe (with name + by when). 2. Decisions I made that should be documented. 3. Things I noticed worth flagging to someone (or to my future self). 4. The thread someone started today that I haven't replied to. 5. A weak signal I shouldn't ignore. </output>

💡

Pro tip: Most "I forgot to follow up" misses happen in the 5pm-7pm gap. A 3-minute capture at end-of-day catches 80% of the forgettable.

Lessons from Today

15/50

✨ What it does

Claude extracts one lesson from today that transfers beyond today, plus a check that you are not overlearning a one-off. You describe the day's surprises, then apply the lesson tomorrow if it still fits.

<task>Extract one transferable lesson from today.</task> <today>[describe today's main interactions / outputs / surprises]</today> <output> 1. One pattern I noticed (could be about my behavior, a person, a process). 2. The transferable lesson: this is generalizable beyond today because... 3. Where it applies tomorrow / this week. 4. The counter-lesson — what I might over-generalize if I'm not careful. 5. The 2-sentence note I'd write to myself in 6 months. </output>

💡

Pro tip: Daily lessons compound. The counter-lesson check ("what would I over-generalize?") prevents the trap of treating a single data point as a pattern.

Promise Check

16/50

✨ What it does

Claude audits promises you made today, maps each to a next action and due time, and flags ones you should not have made. You list the commitments, then renegotiate or schedule the real ones.

<task>Audit promises I made today.</task> <promises>[list things I committed to today — calls, follow-ups, deliveries, intros]</promises> <output> 1. Each promise mapped to a clear next action (verb + by when). 2. Promises I shouldn't have made — explain why and what to do. 3. Promises that need to be renegotiated tomorrow (be honest). 4. Promises already covered by ongoing systems. 5. The single most important promise to keep this week. </output>

💡

Pro tip: Most relational damage comes from forgotten promises, not broken ones. Daily promise capture catches 90% of it before it becomes a trust issue.

Today's Decisions Log

17/50

✨ What it does

Claude logs each decision from today with the real reason, what information you lacked, confidence, and a date to revisit. You list the decisions, then file the log where you will see the revisit date.

<task>Log decisions made today.</task> <decisions>[list decisions made today, big and small]</decisions> <output> For each decision: 1. What was decided. 2. Why (the actual reason, not the rationalization). 3. What information I had vs would have wanted. 4. Confidence level (low / med / high). 5. When to revisit this decision. Then: one decision that should have been delegated. </output>

💡

Pro tip: Decision logs become invaluable in retrospect — pattern-matching your own decision quality over months reveals strengths and traps you can't see in the moment.

What Drained Me Today

18/50

✨ What it does

Claude names the day's biggest drain, whether it is a person, task, or energy state, and how to prevent it next time. You describe today's interactions, then change the slot or the yes that caused it.

<task>Identify the drain in my day.</task> <today>[describe today's main interactions / tasks]</today> <output> 1. The single most draining interaction — who, what, why. 2. The drain pattern: is this person, this task, this energy state? 3. Could the drain have been avoided, reduced, or scheduled differently? 4. Cost: how it affected the rest of the day. 5. One structural change to prevent / contain it next time. </output>

💡

Pro tip: Drains are often patterns — same person, same meeting type, same context switch. Naming them shifts from "bad day" to "bad design."

Yesterday Replay

19/50

✨ What it does

Claude replays yesterday to find a lesson you have not applied yet and the smallest action that would apply it today. You describe yesterday in six bullets, then do that small action.

<task>Replay yesterday to extract what I learned that I haven't applied yet.</task> <yesterday>[describe yesterday in 6 bullets]</yesterday> <output> 1. What I learned yesterday that I haven't acted on. 2. Why it didn't lead to action — friction, forgetting, denial. 3. The smallest action I could take today to apply the lesson. 4. The pattern: am I sitting on learning generally? 5. One follow-up to yesterday that's overdue. </output>

💡

Pro tip: Most lessons get learned and forgotten within 48 hours. The replay catches the lesson before it disappears — combine with the morning routine for compound effect.

Today vs Plan Audit

20/50

✨ What it does

Claude audits the gap between today's plan and what you actually did, including the biggest miss and whether the gap is growing. You list plan versus actual, then adjust tomorrow's plan from the trend.

<task>Audit gap between today's plan and today's actual.</task> <plan>[list what I planned]</plan> <actual>[list what I did]</actual> <output> 1. Plan vs actual: % completion. 2. The biggest miss — why? 3. The biggest unplanned task — what triggered it, was it worth it? 4. Is the plan-actual gap getting bigger or smaller over the last 2 weeks? 5. What this says about how I'm planning vs how I'm working. </output>

💡

Pro tip: Plan vs actual is a forecasting muscle. Track gap % over 30 days — if it's widening, you're overestimating; narrowing, you're underestimating.

Evening Journal

10 prompts

Tonight's Three Questions

21/50

✨ What it does

Claude asks three end-of-day questions one at a time, then turns your answers into sentences for tomorrow, release, and sleep. You answer as today felt [tired / energized / mixed / heavy], then keep the sentences you want.

<task>Ask me the three most useful end-of-day questions.</task> <context> - Today felt: [tired / energized / mixed / heavy] - Tomorrow looks: [packed / open / pivotal] </context> <output> Ask me, one at a time: 1. What was I avoiding today? What's the cost of continuing to avoid it? 2. What's the bravest thing I did today, even if small? 3. What would I forgive myself for tonight if I could? Then synthesize my answers into 3 sentences: one for tomorrow, one to release, one to take into sleep. </output>

💡

Pro tip: The "what was I avoiding?" question is the highest-value journal prompt — it surfaces the actual problem most other tools miss.

What Am I Proud Of Today

22/50

✨ What it does

Claude names one specific thing you can feel proud of today and the win you are dismissing as not real. You paste the messy version of the day, then let yourself keep that pride.

<task>Ground me in what I'm actually proud of today.</task> <today>[describe today in 4-6 bullets, the messy version]</today> <output> 1. The one thing I should let myself feel proud about — be specific. 2. The pride I'm minimizing (the win I'm dismissing as "not real"). 3. What this win says about how I work or who I am. 4. The version of me that did this — describe. 5. A note from future me thanking present me for today. </output>

💡

Pro tip: High performers minimize wins by default. The "pride I'm minimizing" prompt overrides the dismissal habit and reinforces what's working.

Worry Defuse

23/50

✨ What it does

Claude defuses the worry you are taking to bed: what is actually likely, one action item, and a calmer reading. You describe the worry and the worst case, then do the one action or let it wait.

<task>Help me defuse the worry I'm carrying into bed.</task> <worry> - What it is: [describe] - Probability it actually happens: [0-100%] - Worst case if it does: [describe] - What I'd do then: [describe] </worry> <output> 1. Decatastrophize: distinguish what's actually likely from what's worst-case. 2. The action item: is there one move I can make tonight that reduces this worry by half? 3. The action item that requires tomorrow's daylight (so I can release it tonight). 4. The reframe in two sentences I can read before sleep. 5. The "if I wake at 3am" 1-sentence anchor. </output>

💡

Pro tip: Sleep loss usually comes from a worry that has no nighttime action. The "tomorrow daylight" tag lets you release it; the "3am anchor" catches you if it returns.

Gratitude With Depth

24/50

✨ What it does

Claude writes specific, non-generic gratitude from your day, including one person to thank and optional wording for a message. You list people and moments, then send the thank-you if it still feels true.

<task>Help me articulate today's gratitude with depth (not generic).</task> <context> - People I interacted with today: [list] - Moments I noticed: [list] - Things that went right: [list] </context> <output> 1. One person to thank — and the specific thing that earned it. 2. One moment that wasn't obviously good but contained a gift. 3. One privilege I have that I take for granted (be specific). 4. One thing about my work that I'd miss if it ended tomorrow. 5. A 2-sentence note to one person above — that I'll actually send or not. </output>

💡

Pro tip: Generic gratitude ("I'm grateful for my family") doesn't move neurochemistry — specific gratitude does. The 2-sentence message option turns gratitude into relationship.

Identity Check

25/50

✨ What it does

Claude checks the gap between the self you want to become and who you were today, celebrating matches and naming drift without shame. You describe the target traits and today's behaviors, then pick one small close-the-gap move.

<task>Check the gap between who I want to be and who I was today.</task> <context> - The version of me I'm trying to become: [describe in 2-3 traits] - Today's behaviors: [be honest about 3-4] </context> <output> 1. Where today's me matched the target — celebrate it. 2. Where today's me drifted — describe without shame. 3. The smallest behavior change tomorrow to close the gap by 5%. 4. The internal narrative that drove the drift — name it. 5. A line to tell myself tomorrow morning when I'm tempted to drift again. </output>

💡

Pro tip: Identity drift compounds. Daily small checks beat quarterly reckonings — 5% adjustments per day end up bigger than dramatic resets.

Sleep Permission

26/50

✨ What it does

Claude writes an honest permission to stop working tonight, with the real cost of stopping versus pushing on, plus a sentence you can say out loud. You list leftover tasks, then stop when that sentence lands.

<task>Help me actually permission myself to stop working tonight.</task> <state> - Tasks not done today: [list] - The pull to keep going: [describe the urge] - Stakes if I stop now: [be honest — usually less than the urge implies] </state> <output> 1. The cost of stopping now (honestly, not anxiously). 2. The cost of pushing past my real stop time (sleep, judgment, tomorrow's output). 3. The actual emergency tasks (95% won't be). 4. The "tomorrow morning will handle it" list. 5. The permission slip: a single sentence that gives me permission to close the laptop. </output>

💡

Pro tip: High performers struggle to stop more than they struggle to start. The explicit permission slip is unexpectedly powerful — naming the urge defuses it.

Tomorrow's Self, Tonight

27/50

✨ What it does

Claude writes a four-sentence note from tonight-you to tomorrow-you about the lesson, the important thing, and the risk. You name those three, then read the note first thing in the morning.

<task>Write a 4-sentence note from tonight's me to tomorrow's me.</task> <context> - Today's main lesson: [describe] - Tomorrow's most important thing: [name it] - Tomorrow's biggest risk (energy, distraction, drift): [name it] </context> <output> A 4-sentence note that I'll read first thing tomorrow: 1. The one thing I want tomorrow-me to remember. 2. The one thing tomorrow-me should ignore. 3. The one thing tomorrow-me should celebrate before starting. 4. The closing line — short, true, useful. </output>

💡

Pro tip: Future-self letters work because they bypass morning resistance — when tomorrow-you reads it, the framing is already done. Keep it short or it won't get read.

The Hard Thing

28/50

✨ What it does

Claude names the hard thing you have been avoiding, why you avoid it, and the smallest possible next step. You hint at the thing and how long it has sat, then take that small step or schedule it.

<task>Name the hard thing I've been avoiding.</task> <state> - The avoided thing: [hint at it] - How long I've been avoiding it: [duration] - What it costs me to keep avoiding: [list] </state> <output> 1. Name it precisely — vague avoidance stays avoidable. 2. Why I'm avoiding (fear of conflict, identity, outcome, embarrassment). 3. The cost of avoiding for one more week. 4. The smallest possible next step — 5 minutes worth. 5. The single sentence to say (or write) tomorrow that starts it. </output>

💡

Pro tip: Avoidance is rarely about the task — it's about the feeling. Naming the feeling shrinks the task. The "5-minute next step" makes it physically possible to start.

Body Check

29/50

✨ What it does

Claude reads what your body is saying tonight and names the signal under the shoulders, jaw, or stomach, plus one physical reset. You fill in the body, emotion, and mental state, then do the reset before sleep.

<task>Check what my body is saying tonight.</task> <state> - Physical: [shoulders, jaw, stomach, sleep last night] - Emotional: [name 2-3 emotions present] - Mental: [scattered, focused, foggy, anxious, clear] </state> <output> 1. The signal under the symptom — what's the body telling me? 2. One thing to release tonight (a tab, a worry, a tension). 3. Tomorrow morning's first physical move (water, walk, stretch). 4. The conversation I might need to have (with someone or with myself). 5. The 1-sentence "I'm okay" anchor I can read. </output>

💡

Pro tip: Body signals predict burnout 30-60 days before mental signals catch up. Routinizing body check-ins catches it early — most people only check when it's too late.

Five-Year Voice

30/50

✨ What it does

Claude writes a four-sentence message from five-years-from-now you about today's tension, decision, and emotion. You name those three, then keep the advice that loosens tonight's grip.

<task>Channel five-years-from-now me to advise current me about today.</task> <context> - Today's biggest tension: [describe] - Today's biggest decision: [name] - Today's biggest emotion: [name] </context> <output> A 4-sentence message from 5-years-from-now-me to current-me: 1. What 5-year-me wishes I'd known to release tonight. 2. What 5-year-me wishes I'd known to take more seriously. 3. What 5-year-me thinks I'm overestimating right now. 4. The reframe that makes tonight feel smaller and clearer. </output>

💡

Pro tip: The five-year voice is unusually effective at shrinking present worries — most things that feel huge tonight are barely remembered in a year. Use sparingly; it loses power if overused.

These prompts give you the what. Tutorials give you the why.

Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. 300+ tutorials and growing.

Try AI Academy Free

Email & Communication Prep

10 prompts

Three Emails Tonight

31/50

✨ What it does

Claude drafts three ready-to-send emails so tomorrow morning is not a scramble, with when to send and what reply to expect. You list people you owe and decisions to share, then send or schedule the drafts.

<task>Draft the 3 emails I should send tonight so tomorrow morning isn't scrambling.</task> <context> - People I owe replies to: [list] - Decisions I made today that need to be communicated: [list] - Tomorrow's first meeting: [topic + attendees] - Anything I want to set up before 9am: [list] </context> <output> 3 ready-to-send emails. For each: - To: [name] - Subject: [line] - Body: 4-7 sentence draft - The "best time to send" recommendation (now / 7am / right before they wake) - The reply I'm expecting (so I'm not surprised) </output>

💡

Pro tip: The "best time to send" recommendation matters — late-night sends feel anxious to receivers. 7am sends look thoughtful and pre-planned.

Apology / Repair Email

32/50

✨ What it does

Claude drafts a specific apology to [person] that names what happened and the impact, without a blanket sorry or a promise you cannot keep. You fill in what you are and are not apologizing for, then edit and send.

<task>Draft an apology or repair email I've been avoiding.</task> <context> - To: [person] - What happened: [describe] - What I want to repair: [trust, business relationship, friendship] - What I am NOT apologizing for: [be honest with yourself] </context> <output> 1. The actual apology — specific, not blanket. 2. What I understand about the impact on them. 3. What I'll do differently. 4. What I won't do (no over-promising). 5. The send-or-not check: is sending tonight better than waiting? </output>

💡

Pro tip: Blanket apologies ("sorry for everything") feel safer and land weaker. Specific apologies are scarier to write and 5x more effective at actually repairing.

Decline Email

33/50

✨ What it does

Claude drafts a four-sentence decline for [request] that says no clearly, plus replies if they push back. You add the real reason and whether to keep the relationship, then send the no.

<task>Draft a clean decline email for [request].</task> <request> - From: [person] - What they're asking: [describe] - Why I'm saying no: [the real reason] - Relationship to preserve: [yes / no] </request> <output> A 4-sentence decline: 1. Acknowledge their ask specifically. 2. Decline clearly (no hedging language). 3. Reason — brief, honest, no over-explaining. 4. Optional: door-open for the future or a redirect to someone else. Then: the 3 most common follow-up pushbacks and short replies for each. </output>

💡

Pro tip: Over-explaining is the most common decline mistake — it invites negotiation. The 4-sentence template forces brevity, which lands as confidence, not rudeness.

Status Update to Team

34/50

✨ What it does

Claude drafts a short status update for [team / leadership / investor / customer] with one headline, wins, misses, and explicit asks. You list those items, then send the update as is or tighten the headline.

<task>Draft tonight's status update to my team / stakeholder / investor.</task> <context> - Audience: [team / leadership / investor / customer] - Period: [day / week / month] - Wins: [list] - Misses: [list] - Asks: [list] </context> <output> A short update (3-4 paragraphs max): 1. Headline: the one thing they need to know. 2. Wins (specific, with numbers). 3. Misses (with reasoning, no spin). 4. Asks (specific, with deadline). 5. The 1-sentence preview of next period. </output>

💡

Pro tip: Status updates fail when wins are vague and misses are spun. Spin destroys trust faster than misses do — investors and team can read both.

Hard Conversation Outline

35/50

✨ What it does

Claude outlines tomorrow's hard conversation with [person] about [topic]: opening sentence, three points, and what you will hold versus flex. You add the outcome you want, then use the outline in the room.

<task>Outline a hard conversation I need to have tomorrow.</task> <context> - With: [person] - About: [topic] - Outcome I want: [be specific] - What they'll likely push back with: [predict] - Relationship stakes: [list] </context> <output> 1. The opening sentence — clear, no warm-up. 2. The 3 key points, in order. 3. What to ask them after each point. 4. The pre-decided "I won't budge on": list. 5. The pre-decided "I could flex on": list. 6. The exit lines: "let's pause and revisit" and "we don't need to resolve today." </output>

💡

Pro tip: The pre-decided "I won't budge / I could flex" lines are the single most useful prep — they prevent the in-conversation drift where you agree to things you didn't intend to.

Slack DM Cleanup

36/50

✨ What it does

Claude sorts open Slack threads into a one-line reply, a tomorrow tag, archive, or handoff. You paste the threads you owe, then send the quick replies and close the rest.

<task>Clean up my Slack DMs / inbox before bed.</task> <inputs> - Open threads I owe replies on: [list] - Threads I've been avoiding: [list] - Threads I should mark unread / archive: [list] </inputs> <output> For each thread: 1. Quick reply (1-line draft) OR 2. "Tomorrow morning" tag with what I'll say OR 3. Archive — no action needed OR 4. Delegate / forward to someone else. Then: my "inbox-clear sentence" — the closing line that lets me close the app tonight. </output>

💡

Pro tip: Open threads at night cost cognitive load even when ignored. The "tomorrow morning tag" creates a clear action without forcing tonight's reply.

Thank-You Email

37/50

✨ What it does

Claude drafts a three-sentence thank-you to [person] that names [specific action] and [specific impact], plus when to send it. You fill those in, then send the note.

<task>Draft a thank-you email to someone who helped me today.</task> <context> - To: [person] - What they did: [specific action] - Why it mattered: [specific impact] - Relationship: [colleague / mentor / customer / stranger] </context> <output> A 3-sentence thank-you: 1. The specific thing they did (no generic "thanks for everything"). 2. The specific impact it had. 3. The future-tense touch — what you'd like them to know going forward. Then: send-time recommendation (now / morning / wait a day for impact?). </output>

💡

Pro tip: 24-hour gratitude lands better than instant gratitude — it shows you noticed and reflected. The exception is when reciprocity matters: customers, candidates, partners send same-day.

Negotiation Position Email

38/50

✨ What it does

Claude drafts your position email for [deal / ask] with an opening number, why it is fair, and a line if they say no. You add [floor and ceiling] and what they want, then send the version you can stand behind.

<task>Draft my negotiation position email for [deal / ask].</task> <context> - Counterparty: [who] - What I want: [outcome] - What they want: [their outcome if known] - BATNA: my best alternative if this doesn't work - Range: [floor and ceiling] </context> <output> 1. Opening anchor — where I start. 2. The justification — 2 concrete reasons backing the anchor. 3. The signal of flexibility — what I might trade. 4. The deadline / next step. 5. The "what if they say no" line I'd reply with (drafted in advance). </output>

💡

Pro tip: The pre-drafted "if no" reply is the unlock. Most negotiation regret comes from improvised responses to pushback — write yours tonight, not in the moment.

Reply to the One That's Been Sitting

39/50

✨ What it does

Claude drafts the reply you have been avoiding, in an honest version and a softened version, plus a send recommendation. You paste the sitting message, then send the version that is still true.

<task>Draft the reply to the email / message I've been avoiding.</task> <message>[paste the message or describe]</message> <context> - Why I've been avoiding: [be honest] - What they actually need: [their real ask] - What I'd say if I weren't worried: [the honest answer] </context> <output> 1. The 4-sentence honest reply. 2. The kinder version of the same reply (sometimes the honest one needs softening, sometimes it doesn't). 3. The send recommendation: now / tomorrow morning / talk-in-person instead. 4. The avoidance cost: what's it costing me to delay further? </output>

💡

Pro tip: The "kinder version" check protects you from sending honest-but-blunt. Most avoided emails get sent eventually — Claude lets you send sooner and better.

End-of-Day Slack Closer

40/50

✨ What it does

Claude drafts a two- or three-line Slack closer that says you are offline, what is handed off, and when you are reachable. You list what is still open, then post the closer and actually stop.

<task>Draft my end-of-day Slack closer (the message that signals I'm done).</task> <context> - Team norm: [explicit signing off or just silent] - What's open: [items others might need] - Tomorrow's first commitment: [time / topic] </context> <output> A 2-3 line Slack message: 1. Quick "I'm offline" line. 2. What's handed off and to whom (if anything). 3. When I'm back tomorrow and what's first. Plus: my optional "if you need me urgently, here's how" line (or the boundary that I won't be reachable). </output>

💡

Pro tip: Explicit sign-offs work better than going silent — people don't wonder if you're working, and you don't feel guilty for closing the app.

Decision & Anxiety Reframes

10 prompts

Decision in 2 Sentences

41/50

✨ What it does

Claude restates the decision you are sleeping on in two sentences and checks whether it is easy to undo. You describe [option A vs option B] and the fear, then decide tonight or let it wait with a clear reason.

<task>Reframe a decision I'm sleeping on into 2 sentences so I can release it tonight.</task> <decision> - The decision: [describe] - What I'm caught between: [option A vs option B] - What I'm afraid of: [be honest] - The deadline: [actual or self-imposed] </decision> <output> 1. The decision in 2 sentences — what's actually being decided. 2. The reversibility check — Type 1 (irreversible) or Type 2 (reversible)? 3. If Type 2: just pick — explain why I'm overthinking. 4. If Type 1: what's the missing information that'd make this clearer? 5. The deadline for tonight: just enough to release the worry without forcing a yes. </output>

💡

Pro tip: Most decisions you sleep on are Type 2 (reversible). The reversibility check usually reveals you can just pick — most overthinking is wasted on undoing.

3am Worry Disarm

42/50

✨ What it does

Claude disarms the bedtime loop: names the replay, the real odds, what you can influence, and a morning action written in advance. You describe what is looping, then use the morning action instead of rehearsing it.

<task>Disarm the worry I'm bringing into bed before it wakes me at 3am.</task> <worry>[describe what's looping]</worry> <output> 1. Name the loop: what scenario is replaying? 2. Reality check: actual probability of the catastrophic outcome (be honest, often 5-15%). 3. Locus of control: what's mine to influence vs not. 4. Action available tonight: yes or no. 5. The anchor sentence to read if I wake at 3am. 6. The morning's first action — written down so 3am-me can release it. </output>

💡

Pro tip: The "morning action written down" trick stops the loop because the brain releases what it knows is captured. The 3am replay is the brain re-checking that the task isn't lost.

Either-Or Tiebreaker

43/50

✨ What it does

Claude breaks a tie between Option A and Option B using cost of being wrong, later regret, and how easy each is to undo. You describe both options and your gut pick, then choose or ask for the one missing fact.

<task>Break the tie between Option A and Option B.</task> <context> - Option A: [describe] - Option B: [describe] - What I'd do if I had to choose now (gut): [honest answer] - What's making me hesitate: [name it] </context> <output> 1. Cost of getting it wrong with A vs wrong with B. 2. Information that'd change the answer — is it gettable? 3. The "regret in 5 years" test: which would I regret more? 4. The reversibility test: which can I undo? 5. Recommendation: pick now, or wait — explain why. </output>

💡

Pro tip: Most "I can't decide" stalls are between options that are 49/51. Pick the 51 and move — the 2% downside is dwarfed by the cost of standing still.

The One Decision Worth Sleeping On

44/50

✨ What it does

Claude sorts [list pending decisions] into decide-tonight versus sleep-on, based on whether tomorrow brings new information and whether you can undo it. You apply the split: decide the first pile and leave the rest.

<task>Filter today's decisions: which ones to sleep on vs which to just decide tonight.</task> <decisions>[list pending decisions]</decisions> <output> For each decision: 1. Is this Type 1 (irreversible) or Type 2 (reversible)? 2. Sleep-on test: would I actually have more information tomorrow? 3. Just-decide test: would I be the same person making this decision tomorrow? 4. Recommendation: decide tonight OR sleep on it (and only the ones that pass the sleep-on test). Then: tonight's "I just decided" list and tomorrow's "still thinking" list. </output>

💡

Pro tip: Most "I'll sleep on it" is procrastination disguised as wisdom. Only sleep on it if more information actually arrives by morning — otherwise just decide.

Catastrophizing Catch

45/50

✨ What it does

Claude catches a catastrophic story, names the distortion, and offers a more likely reading of the same facts. You describe the story and the evidence you are citing, then keep the reading that matches reality.

<task>Help me catch and reframe a catastrophic narrative.</task> <narrative> - The story I'm telling myself: [describe] - The worst case in my head: [describe] - The evidence I'm citing for this fear: [list] </narrative> <output> 1. Name the cognitive distortion (catastrophizing, mind-reading, fortune-telling, all-or-nothing). 2. The actual probability of the catastrophic outcome. 3. What evidence contradicts the story — list it. 4. The neutral version of the same situation. 5. The alternative reading: what's a more useful story I could tell about this tonight? </output>

💡

Pro tip: Naming the cognitive distortion ("oh, this is catastrophizing") is itself the unlock. The brain can't maintain the loop as easily once it's named.

Comparison Trap Exit

46/50

✨ What it does

Claude pulls you out of a comparison by naming what you are not seeing about them and what you are minimizing about yourself. You name who you are comparing to, then drop the comparison or use it as one data point.

<task>Pull me out of the comparison trap.</task> <state> - Who I'm comparing myself to: [name or describe] - What they have / are: [the thing I'm fixating on] - What I'm minimizing about my own situation: [be honest] </state> <output> 1. The fair comparison would be — describe. 2. What I'm not seeing about their situation (cost, struggle, trade-off). 3. What I'm not seeing about my own situation (advantages I'm ignoring). 4. The internal benchmark I should use instead. 5. A single sentence to break the loop tonight. </output>

💡

Pro tip: Comparison runs on incomplete information — you see their highlights, you live your behind-the-scenes. Naming the hidden costs of their situation usually breaks the spiral.

Resistance Read

47/50

✨ What it does

Claude reads the resistance toward [thing I should do tomorrow] by separating the surface excuse from the fear underneath. You describe the resistance, then do a smaller version or name the real danger if it is real.

<task>Read the resistance I'm feeling toward [thing I should do tomorrow].</task> <context> - The thing: [describe] - The resistance: [what I'm telling myself] - What I'm afraid of underneath: [guess] </context> <output> 1. The surface excuse vs the actual fear. 2. Is this resistance protecting me from real danger or just discomfort? 3. The smallest scary step I could take tomorrow that'd defuse 80% of the resistance. 4. The evidence I'm strong enough to handle the discomfort. 5. The 1-sentence permission to be uncomfortable tomorrow and do it anyway. </output>

💡

Pro tip: Resistance is usually proportional to importance — the harder you resist something, the more it likely matters. The "smallest scary step" makes it physically possible to start.

Imposter Disarm

48/50

✨ What it does

Claude disarms imposter feelings before [meeting / talk / pitch] with a list of evidence you already have and what you actually add. You list why you feel like a fraud and what qualifies you, then reread the evidence list tomorrow.

<task>Disarm imposter feelings before tomorrow's [meeting / talk / pitch].</task> <context> - The event: [describe] - Why I feel like a fraud: [be honest] - What I've actually done that qualifies me: [list] - Who I'm worried will see through me: [describe] </context> <output> 1. Evidence list: 3-5 things I've actually accomplished that qualify me to be in this room. 2. The room reality: everyone there has imposter moments — name 2-3 likely. 3. The shift: from "do I deserve to be here?" to "what value can I add?" 4. The pre-event grounding sentence. 5. The post-event "I survived" prompt for tomorrow night. </output>

💡

Pro tip: Imposter syndrome shrinks when you shift from "do I deserve to be here?" to "what value can I add?" The first is identity; the second is action.

Conflict Pre-Frame

49/50

✨ What it does

Claude pre-frames tomorrow's conflict with [person] by writing their strongest case, your core point, and a line that stops escalation. You add [topic] and where you might be rigid, then walk in with that frame.

<task>Pre-frame tomorrow's conflict so I show up grounded.</task> <context> - With: [person] - What we're conflicting about: [topic] - Where I think they're coming from: [steelman it] - Where I'm probably wrong or rigid: [be honest] </context> <output> 1. Steelman of their position (be generous). 2. My core point — distilled to 2 sentences. 3. What I'm willing to be wrong about going in. 4. The win-for-both outcome I could pitch. 5. The sentence to say if I feel myself escalating tomorrow. </output>

💡

Pro tip: Pre-deciding what you're willing to be wrong about is the most underrated conflict prep. You enter open, which de-escalates faster than any technique.

Release the Day

50/50

✨ What it does

Claude writes a four-line release ritual to read before sleep: what today asked, what you put down, what can wait, and permission to rest. You name the heaviest and lightest things, then read the four lines and stop.

<task>Give me a 4-sentence "release the day" ritual.</task> <state> - Heaviest thing from today: [name it] - Lightest thing from today: [name it] - Anything left undone: [list briefly] </state> <output> A 4-line release ritual to read before sleep: 1. The acknowledgment — what today asked of me. 2. The release — what I'm not carrying into tomorrow. 3. The trust — what's handled or will be tomorrow. 4. The closing — short, true, mine. </output>

💡

Pro tip: Sleep quality correlates more with how you close the day than how you start it. A 60-second release ritual every night does more for sleep than any morning routine.

Free tool

Prompt Optimizer

Turn a rough idea into a structured, professional AI prompt.

Try it free →

Frequently Asked Questions

5-10 minutes total. Most prompts produce useful output in 30-90 seconds. The slowest part is pasting context — calendar, tasks, what's on your mind. With a Claude Project that holds your context, it drops to under 5 minutes.
Claude does what a journal can't: it asks you targeted questions, surfaces patterns, and produces structured output (tomorrow's plan, draft emails, decision frames). Journals capture; Claude processes. Use both if you like — they're complementary.
Yes. Most evening prompts are short enough to type or voice-dictate. The "Tomorrow at a Glance" and "Three Emails Tonight" are particularly mobile-friendly — they're what most people use from bed.
Yes — create one Project called "Evening Routine" with your context (calendar style, role, current focus, sleep struggles if relevant). Then every prompt runs against that context automatically. Cuts setup time by 80%.
Specifics in, specifics out. Replace placeholders with actual details (real meeting names, real worry text, real numbers). Generic input produces generic output — Claude is only as good as what you give it to work with.
Not by default — each conversation starts fresh. Use Claude Projects to save persistent context, or copy a 2-sentence "yesterday's key things" into tonight's session. Most people develop a 30-second start-of-evening note that catches Claude up.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.