メニュー
Determine how much thinking to invest in a decision. Maps stakes, reversibility, and diminishing returns to effort levels (minimum, standard, maximum).
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Route any input through a branching question tree to narrow down the optimal response strategy before writing. Two stages — PERCEIVE (classify input) then ACT (select response). Covers all prompt types.
Generate exhaustive guesses about user input using ALL search methods with coverage tracking. Guessing is SEARCH through possibility space. Tracks space created vs space covered to ensure comprehensive exploration.
Systematically evaluate and select from a set of guesses, options, or possibilities. Combines ARAW analysis with prioritization to determine which guesses are strong, weak, actionable, or eliminable.
Extract hidden assumptions from any content. Surfaces what must be true for claims to hold, enabling deeper analysis.
Sub-orchestrator for analytical requests. Routes to decomposition, systems analysis, comparison, risk assessment, or synthesis based on what kind of analysis is needed.
Assume Right - Deep recursive rightness search. For every claim, assume it's right — find what must follow, then assume THOSE implications are right too. Recurse until bedrock. Track every claim found.
| name | ecal - Effort Calibration |
| description | Determine how much thinking to invest in a decision. Maps stakes, reversibility, and diminishing returns to effort levels (minimum, standard, maximum). |
| output | {"format":"prose"} |
Input: $ARGUMENTS
Before executing, identify which interpretation matches the user's input:
Interpretation 1 — Calibrate effort for a pending decision: The user has a specific decision ahead and wants to know how much thinking, research, and process it actually warrants before they begin. Interpretation 2 — Escape analysis paralysis: The user is already deep in deliberation on something and suspects they are over-investing — they need a way to cut through and commit. Interpretation 3 — Build an effort-calibration habit: The user wants to internalize the framework itself so they can routinely right-size their decision-making process across many future choices.
If ambiguous, ask: "I can help with calibrating effort for a specific decision, escaping analysis paralysis on something you're stuck on, or building the habit of right-sizing effort — which fits?" If clear from context, proceed with the matching interpretation.
State the decision in one sentence. If you cannot, the problem is upstream — use /sid or /reframe first.
Answer these three questions. Do not deliberate. Use your first instinct.
Q1: "If I flip a coin right now, would the outcome be roughly acceptable?"
Q2: "If I get this wrong, will I still care in 6 months?"
Q3: "Can I easily reverse or adjust this decision after making it?"
Override rule: If any answer took more than 15 seconds of hesitation, bump up one effort level. Hesitation itself is information that the decision is more complex than it appears.
Meta-guard: The triage itself must take under 60 seconds. If you catch yourself deliberating about the triage questions, pick Section B (Standard) and move on. The recalibration check in Section D exists precisely for this case.
When to use: Trivial stakes, freely reversible, or coin-flip-acceptable decisions.
Total time budget: 2 minutes maximum.
Procedure:
A1. State the decision in one sentence.
A2. List at most 3 options. (If you cannot think of 3, that is fine. If you can think of more than 3, stop at 3 -- you are over-investing.)
A3. Pick the one that feels right. Do not justify it. Do not list pros and cons. Gut feeling is sufficient for coin-flip-acceptable decisions.
A4. Act immediately. Do not revisit.
Hard constraint: If you reach the 2-minute mark without having decided, pick whichever option you were most recently considering and execute it. The cost of continued deliberation now exceeds the cost of any wrong answer.
When to use: Moderate stakes, reversible with some cost, "I would care if I got this wrong but it would not be catastrophic."
Total time budget: 5-30 minutes, depending on scope. Set a timer. When the timer expires, decide with whatever you have.
Procedure:
B1. State the decision in one sentence and identify the single most important criterion (the one factor that, if satisfied, makes the decision good enough regardless of other factors).
B2. List 3-5 options. If you have fewer than 3, spend 2 minutes generating more. If you have more than 5, eliminate the weakest until you are at 5. Do not spend more than 3 minutes on this step.
B3. Evaluate each option against your single most important criterion only. Score each option as STRONG / ADEQUATE / WEAK on that criterion. Eliminate all WEAK options.
B4. Among remaining options, check for dealbreakers -- any single factor (not your primary criterion) that would make an option unacceptable. Eliminate any option with a dealbreaker.
B5. From surviving options, pick the one with the most upside or the one you can start executing fastest (speed of execution is a tiebreaker, not a primary criterion).
B6. State your decision. Write one sentence on why. Execute.
B7. Schedule a check-in. For any Standard-effort decision, note when you will assess whether the decision was correct (e.g., "Revisit in 2 weeks"). This is your safety net for cases where you should have used Section C.
Hard constraint: If you are past 30 minutes, you have left Section B territory. Either commit now or admit this is a Section C decision and restart there.
When to use: High stakes, practically irreversible, "getting this wrong would haunt me."
Total time budget: 1-8 hours of active analysis, spread across at least 2 sessions with a break between them (sleep on it if the timeline allows). Set total time budget at the start; do not extend it by more than 25%.
Procedure:
C1. Frame: State the decision, the deadline, the key stakeholders, and the 2-3 most important criteria.
C2. Enumerate: List all plausible options (aim for 5-10). Include at least one unconventional option and the option of "do nothing / delay."
C3. Research: For each option, identify the 1-3 facts that would most change your assessment if they turned out differently than you assume. Spend your research time on those specific facts, not on general background. Set a research time limit equal to 40% of your total time budget.
C4. Evaluate: Score each option against your 2-3 criteria. Use a simple scale (1-5 or Low/Medium/High). Do not use weighted scoring formulas -- they create false precision. Instead, look at the scores and ask: "Does any option dominate (better on all criteria)? Does any option have a fatal flaw?"
C5. Stress test the leader: Take your current frontrunner and actively argue against it for 5 minutes. What is the strongest case for a different option? If the counter-argument is compelling, switch to a comparison between the top 2 only.
C6. Pre-mortem: Assume you chose the frontrunner and it failed badly. Write 2-3 sentences describing the most likely failure scenario. Then ask: "Can I mitigate this failure mode?" If yes, note the mitigation and proceed. If no, seriously consider the second-place option.
C7. Decide: Commit. Write your decision, the top reason for it, and the one thing that would make you reverse it (your "reversal trigger").
C8. Act: Begin execution within 24 hours. A decision that is not acted on decays rapidly.
Hard constraint: You must reach C7 within your stated time budget plus 25% buffer. If you are not there, apply the Emergency Close: pick the option that is best on your single most important criterion and commit.
When to run: At the natural midpoint of any analysis, or when you feel a vague sense of "this is taking longer than it should" or "this feels too casual for what's at stake."
Procedure (complete in under 2 minutes):
D1. Restate what you are deciding in one sentence.
D2. Restate your current effort level (A / B / C).
D3. Ask: "Given what I have learned SO FAR in this analysis, is this effort level still correct?"
D4. Ask: "Am I actually making progress, or am I circling?" Circling means: you are generating new considerations but none of them are changing your leaning. If circling, skip to the decision step of your current section.
D5. Ask: "Would taking a small action right now teach me more than continued analysis?" If yes, pause the analysis, take the action, and resume with new information.
Hard constraint: The recalibration check must complete in 2 minutes. If it does not, your answer is "I am over-investing" -- drop one section level immediately.
When to use: You realize you have been thinking about this for far longer than the decision warrants. You feel stuck, overwhelmed, or unable to choose.
Procedure:
E1. Stop all analysis immediately. Do not finish the thought you are in the middle of.
E2. Write down your current top 2 options (no more). If you cannot identify 2, write down the last option you were considering and "the opposite of that."
E3. For each option, complete this sentence: "If I commit to this right now and never look back, the most likely outcome is ______." Write the most likely outcome, not the worst case.
E4. Pick the option with the better most-likely outcome. If they are approximately equal, pick the one that is faster to execute.
E5. Commit. Announce your decision out loud, in writing, or to another person. Making the commitment external makes it real.
E6. Set a "no-revisit" window: you are not allowed to reconsider this decision for a specified period (at least 48 hours for minor decisions, at least 2 weeks for major ones), UNLESS your pre-defined reversal trigger fires.
E7. Forgive the sunk cost. The time you spent over-analyzing is gone. It does not mean the decision needs to be proportionally momentous. A $2 decision that you accidentally spent 3 hours on is still a $2 decision.
STEP 0 — TRIAGE (< 60 seconds)
Q1: Coin flip acceptable? → YES → Section A
Q2: Care in 6 months? → NO → Section A or B
Q3: Easily reversible? → YES → Section B | NO → Section C
Hesitated on any Q? → Bump up one level
Triage taking > 60 sec? → Default to Section B
SECTION A — MINIMUM (< 2 min)
Name it → 3 options max → Pick by gut → Act → Do not revisit
SECTION B — STANDARD (5-30 min, set timer)
Name it + 1 key criterion → 3-5 options → Score on key criterion →
Eliminate dealbreakers → Pick winner → State why in 1 sentence →
Schedule check-in → Execute
SECTION C — MAXIMUM (1-8 hrs, across 2+ sessions)
Frame → Enumerate (5-10 options) → Research (40% of budget) →
Score (2-3 criteria) → Stress test leader → Pre-mortem →
Decide with reversal trigger → Execute within 24 hours
SECTION D — RECALIBRATE (< 2 min, run at midpoint)
Restate decision → Check effort level → Progress or circling? →
Would acting teach more than thinking? → Adjust or continue
SECTION E — ESCAPE PARALYSIS
Stop → Top 2 options → Most likely outcome of each →
Pick better one → Commit externally → No-revisit window → Forgive sunk cost
The triage (Step 0) determines effort level. It is not a substitute for the analysis itself. Do not answer Q1-Q3 and then declare the decision made (unless Q1 is YES, in which case Section A is the analysis and it takes 2 minutes).
If you move from Section B to Section C mid-analysis, you must actually allocate more time. Running Section C steps in a Section B time budget produces the worst of both worlds: the overhead of a thorough process without the time to do it properly. Set a new timer when you upgrade.
The question is not "how anxious does this decision make me?" The question is "what happens if I get it wrong?" Anxiety is not evidence of high stakes. A phobic person may be terrified of a trivially-staked decision. Use the triage questions, not your emotional temperature.
The recalibration check (Section D) is most valuable precisely when you feel you are "almost done" -- because that feeling often means you are circling rather than converging. Run the check. It takes 2 minutes.
"How much effort should I spend deciding how much effort to spend?" The answer is: the triage takes under 60 seconds. That is the effort you spend. This procedure was designed to be self-limiting at the meta level. Do not recurse.
The triage gives you a starting point. It is deliberately rough. The Recalibration Check exists because roughly 20-30% of initial assignments are wrong. Always run Section D at least once during any Section B or C analysis.
Generating a 15th option is not analysis. In Section B, you are capped at 5 options. In Section C, you are capped at 10. These caps are not arbitrary -- they reflect the empirical finding that decision quality does not improve beyond 5-7 options and often degrades due to choice overload.
Override the procedure output and escalate to a higher effort level when:
Override the procedure output and de-escalate to a lower effort level when:
Context: You are building a new feature that needs persistent storage. Your team already uses PostgreSQL. There are 3 viable options: PostgreSQL, a new NoSQL database, or a managed cloud service.
Step 0 -- Triage:
Recalibration at midpoint (Section D): After 45 minutes of analysis, you realize: the data model is simple (key-value lookups), the scale is modest (< 1M records), and your team already knows PostgreSQL deeply. The "right" answer was obvious before you started -- use PostgreSQL.
D3: "Is this effort level still correct?" NO -- this is actually a Section A decision. The team's existing expertise and the simple data model make this a non-decision.
Outcome: Drop to Section A. Use PostgreSQL. Total time: 50 minutes (45 wasted, 5 useful). Lesson learned: for future decisions, check "is there an obvious default?" as a pre-triage step.
Context: A customer is upset about a delayed shipment. You need to decide what to offer: apology only, partial refund, full refund, or replacement with expedited shipping.
Step 0 -- Triage:
Section A execution:
Total time: 90 seconds. Appropriate for the stakes.
Context: You have one offer in hand. The role is interesting but requires relocating. The salary is 15% above your current. You have 5 days to decide.
Step 0 -- Triage:
Section C execution (abbreviated):
Total time: ~6 hours across 3 days (research day, reflection day, decision day). Appropriate for the stakes.
This procedure sits upstream of all other analytical procedures. Before following any decision framework, problem-solving method, or analysis template, run Step 0 of this procedure to determine how much of that framework you actually need. Most frameworks are Section C tools being applied to Section A problems.