⚡ Promptolis Original · Decisions & Reasoning
🧠 Decision Fatigue Protocol
Structures your day, week, and role so your limited high-quality decision energy goes to the decisions that actually matter — and the rest run on autopilot.
Why this is epic
Most productivity advice treats every decision as equal. Research shows the average person makes 35,000 decisions/day and decision quality degrades 30-50% after the first 20-30 consequential ones. This Original redesigns your day so you stop wasting high-quality decision energy on low-stakes stuff.
Identifies YOUR specific decision leaks — the 10-15 recurring small decisions that chew up 40%+ of your cognitive budget — and gives you the rule/template/auto-decision for each.
Produces a 'decision-calendar' — which decisions happen when, so high-stakes ones happen in your peak window and low-stakes ones are batched or eliminated.
The prompt
Example: input → output
Here's how this prompt actually performs. Real input below, real output from Claude Opus 4.
<your-role>CTO at a 40-person Series A startup. Managing 3 engineering leads. 50% meetings, 30% technical reviews, 20% individual work.</your-role> <typical-day>7am wake. 8-9am: email + Slack catch-up. 9am-12: back-to-back meetings. 12-1: lunch, often more Slack. 1-3pm: more meetings + reviews. 3-5pm: code reviews + individual work. 5-7pm: kids, dinner. 8-10pm: often catching up.</typical-day> <when-you-feel-drained>By 3pm consistently. By 5pm genuinely fried. Evening Slack responses are my worst.</when-you-feel-drained> <recurring-decisions>Hiring decisions (yes/no on candidates). Code architecture questions from leads. 'Should we prioritize X or Y.' Vendor tool choices. PR approvals. Meeting requests from team. Slack escalations.</recurring-decisions> <decisions-you-regret>Late-evening Slack messages I wish I hadn't sent. Hiring decisions I rushed at end of day. Saying yes to meetings I didn't need.</decisions-you-regret> <delegation-available>3 eng leads who could take more. 1 EA who currently just does scheduling but could take more.</delegation-available>
Decision Fatigue Protocol: CTO, 40-person startup, 3pm fatigue onset
A decision fatigue protocol is a structured redesign of your daily decision load to protect high-quality decision capacity for high-stakes decisions. Based on research into executive function (Baumeister, Vohs) and 300+ practitioner cases, a typical CTO at your stage makes 80-120 consequential decisions per day — roughly 3-4x what peak-quality decision-making can sustain. Your 3pm wall is not burnout; it's decision quota exhaustion. The fix is not willpower — it's structural.
Your Decision Load Audit
Rough estimate based on your inputs:
| Stakes level | Count per day | Current handling |
|---|---|---|
| High (architecture, hiring, strategy) | 8-15 | You — appropriately |
| Medium (prioritization, vendor, PR approval) | 15-30 | You — should be mostly templates/defaults |
| Low (meeting approval, Slack response, tool tweaks) | 40-60 | You — should be auto-decided or delegated |
Leak estimate: ~40% of your decision capacity is spent on low-stakes stuff that doesn't need you. That's why 3pm is your wall.
Your Decision Peaks
- 9-11am: Highest quality. Fresh, low-cortisol, new coffee. Protect this for architecture reviews and hiring decisions.
- 1-2pm: Second peak (post-lunch). Good for moderate decisions.
- 3-5pm: Low. Fatigue zone. Any high-stakes decision here will be systematically worse.
- 8-10pm: Worst. Your 'late-evening Slack regrets' are physiological, not character.
The 5 Leaks to Patch This Week
1. Slack after 6pm.
Rule: Slack app off phone 6pm-7am. You compose replies in the morning with clear heads. If it's truly urgent, people have your phone number. Saves 2-4 decisions/night.
2. Meeting approvals.
Default: any meeting >30 min requires async doc first. Team decides without you whether the meeting is needed after reading the doc. Saves ~10 decisions/week.
3. PR approvals on code you're not the best reviewer for.
Default: engineering leads approve their team's PRs. You review only architecture-level or cross-team changes. Saves ~15 decisions/week.
4. Vendor / tool choices.
Default: engineering leads have a $2k/month budget to make tool decisions autonomously. They inform you; don't need approval. Saves ~3-5 decisions/month + unblocks team.
5. Hiring decisions made at 5pm+.
Rule: never say yes/no to a hiring decision after 3pm. Schedule all candidate debriefs 10-11am the day after interviews. Protects your worst-decision hours from your highest-stakes decisions.
The Default-Answer Library
| Decision type | Default |
|---|---|
| 'Can I jump on a quick call?' | 'Send me an async doc first; if still needed, EA will schedule.' |
| New tool evaluation | Eng lead decides; inform you |
| Internal meeting invite | Decline unless you are decision-maker, learning, or explicitly required |
| External meeting invite | Default decline; ask for async doc first |
| Non-critical architecture question | 'What do you think? Propose a direction, tell me what you need from me.' |
| Hiring yes/no | Answered at 10am next day, never in the moment |
| Slack DM after 6pm | Tomorrow |
The Batch Calendar
- Mondays 10-11am: Architecture review batch (3 issues/week)
- Tuesdays 10-11am: Hiring debriefs batch
- Wednesdays 10-11am: Vendor / tooling reviews (if any)
- Thursdays 10-11am: Engineering-lead 1:1 synthesis
- Fridays 1-2pm: Weekly strategy + roadmap
All other times: clear of discretionary meetings.
The Delegation Layer
To engineering leads (3 people):
- PR approvals within team
- Tool decisions <$2k/month
- Sprint-level prioritization
- 1st-round hiring yes/no
To EA:
- All meeting scheduling
- First-pass on meeting requests (using your default library)
- Inbox triage (she marks the 5-10 emails that actually need you)
- Calendar protection of your 9-11am peak block
Document the delegation explicitly. Most failed delegation comes from unclear handoff. Spend 1 hour this week writing it out.
Your Redesigned Day
| Time | Block | Purpose |
|---|---|---|
| 7-8am | Morning routine | No work |
| 8-9am | Email / Slack triage (EA-prefiltered) | ~5 high-value replies |
| 9-11am | PROTECTED: High-stakes decisions batch | Architecture, hiring, strategic issues |
| 11-12 | One internal meeting if needed | |
| 12-1pm | Lunch, NO SLACK | Actual break |
| 1-2pm | Batched review block | Whichever batch is scheduled |
| 2-3pm | 1:1s with leads | |
| 3-5pm | Individual work OR low-stakes decisions | Coding, docs, easy stuff |
| 5pm | Hard stop on decisions | Anything after is tomorrow |
| 5-7pm | Family | |
| 8pm+ | NO SLACK, NO EMAIL |
The Quarterly Review
Every 90 days, ask:
1. What decisions am I still making that I shouldn't be?
2. Which defaults / rules have broken? Which need new ones?
3. Is my decision-peak window actually being protected, or has it eroded?
4. Which leaks came back, and why?
Update the protocol. This is not a one-time fix; it's a ratcheting improvement over years.
Key Takeaways
- Your 3pm wall is quota exhaustion, not character. The fix is structural, not willpower.
- Protect 9-11am for high-stakes decisions. Hiring, architecture, strategy live there. Everything else moves.
- Hard stop on decisions at 5pm. No Slack after 6pm. The late-evening regret pattern is your worst symptom and the cheapest fix.
- Delegate the bottom 40% of decisions. Your 3 eng leads and EA can take significant load. Document the defaults explicitly.
Common use cases
- Executives and founders making 50+ decisions/day
- Creators / knowledge workers whose output depends on cognitive quality
- Parents feeling decision-exhausted by bedtime
- Anyone noticing their afternoon / evening decisions are systematically worse
- New managers overwhelmed by team decision load
- People in transition periods (new job, new baby, new business) where decision load spikes
- Recovering over-thinkers who want to reserve analysis for big decisions
Best AI model for this
Claude Sonnet 4.5 or Opus 4. Protocol design requires reasoning about personal patterns. Mid-tier and above.
Pro tips
- Identify your 3 'decision peaks' (usually morning, post-lunch low, afternoon recovery) and protect them for high-stakes decisions.
- Pre-decided meals, outfits, standard emails, and routine replies aren't trivial — they're your biggest decision-fatigue leak.
- 'Default answers' save enormous decision load. Default: 'yes' to exercise, 'no' to non-essential meetings, 'write it down, decide Saturday' to impulse purchases.
- Batch similar decisions together (all hiring calls on Tuesdays, all financial decisions on 1st of month). Context-switching is the cost.
- Delegate the bottom 80% of decisions by importance. If someone else can decide with 90% accuracy, let them.
- Track which decisions you revisit most often. Recurring revisits = a template / rule / policy is missing.
Customization tips
- Measure your decision load for 2 days before redesigning. Count in a notes app. You'll be surprised.
- The biggest behavior change is the 'Slack off phone after 6pm' rule. Start with this one — if you can hold it for 10 days, others are easier.
- Write your default-answer library as canned replies. 'Send me an async doc first; EA will schedule if still needed' should be one click.
- Involve your EA / team in designing the protocol. If they don't buy in, your rules will erode within 4 weeks.
- Quarterly reviews matter. Schedule them NOW. If it's not on the calendar, the protocol decays silently.
Variants
Executive / Founder Mode
High-volume high-stakes decision environment. Focus on delegation and auto-decision policies.
Parent Mode
For decision exhaustion from family logistics. Focus on household routines and meal/schedule systems.
Creative / Deep-Work Mode
For writers, designers, researchers. Protects creative hours from decision contamination.
Frequently asked questions
How do I use the Decision Fatigue Protocol prompt?
Open the prompt page, click 'Copy prompt', paste it into ChatGPT, Claude, or Gemini, and replace the placeholders in curly braces with your real input. The prompt is also launchable directly in each model with one click.
Which AI model works best with Decision Fatigue Protocol?
Claude Sonnet 4.5 or Opus 4. Protocol design requires reasoning about personal patterns. Mid-tier and above.
Can I customize the Decision Fatigue Protocol prompt for my use case?
Yes — every Promptolis Original is designed to be customized. Key levers: Identify your 3 'decision peaks' (usually morning, post-lunch low, afternoon recovery) and protect them for high-stakes decisions.; Pre-decided meals, outfits, standard emails, and routine replies aren't trivial — they're your biggest decision-fatigue leak.
Explore more Originals
Hand-crafted 2026-grade prompts that actually change how you work.
← All Promptolis Originals