القائمة
Sub-orchestrator for decisions. Routes to comparison, ARAW testing, dominance analysis, or goal clarification based on what the user knows and the stakes involved.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
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 | decide - Make a Decision |
| description | Sub-orchestrator for decisions. Routes to comparison, ARAW testing, dominance analysis, or goal clarification based on what the user knows and the stakes involved. |
| output | {"format":"prose"} |
Input: $ARGUMENTS
Before executing, identify which interpretation matches the user's input:
Interpretation 1 — Binary choice: The user faces a yes/no decision ("Should I X?") where the implicit alternative is "don't X." Interpretation 2 — Multi-option comparison: The user has named options ("X or Y or Z?") and needs help choosing between them. Interpretation 3 — Open decision: The user knows they need to decide something but hasn't identified the options yet ("What should I do about X?").
If ambiguous, ask: "Do you have specific options to choose between, or do you need help identifying what the options are?" If clear from context, proceed with the matching interpretation.
Every decision is a bet on the future. The question is never "which is right?" but "which has the best expected outcome given what I know and what I don't know?" Decisions under uncertainty require different tools than decisions under certainty.
"Should I X?" always has a hidden comparator. The alternative to doing X is not "nothing" — it's whatever the user would do instead. Make the comparator explicit. "Should I quit my job?" compares quitting to staying, but also to other changes that might address the same underlying need.
Criteria must be explicit before comparison. "Which is better?" is unanswerable without knowing "better for what?" If the user can't state criteria, they need /want first to clarify what they're optimizing for.
Reversibility determines depth. Reversible decisions (which framework, which restaurant) need light analysis. Irreversible decisions (quit job, major investment, move countries) need deep analysis with pre-mortem and failure anticipation.
The best option is often "none of the above." If all named options feel wrong, the framing is wrong. Look for the option the user hasn't considered — the reframe, the hybrid, the "wait and gather more information."
Foreclosures matter as much as gains. Every option gives something up. The user often focuses on what they get and ignores what they lose. Make foreclosures explicit for every option.
Not all decisions are expected-value calculations. Some decisions are about identity, commitments, or values. If the decision is identity-driven, name the commitment and test whether it's genuine, not whether it maximizes expected utility.
What is being decided? State the options if visible. If only one option is stated ("Should I X?"), the implicit second option is "don't X."
Before finalizing, consider whether supplementary analysis would help:
| Situation | Also invoke |
|---|---|
| Obvious bad outcomes being ignored | → /obo (obvious bad outcomes) |
| Obvious good outcomes being missed | → /ogo (obvious good outcomes) |
| Decision involves ethical dimensions | → /eth (ethics analysis) |
| Decision involves safety risks | → /saf (safety analysis) |
| Decision could be reframed | → /iaw (in another way) |
| User might be self-deceiving | → /sdc (self-deception check) |
| Decision involves future predictions | → /fut (future analysis) |
| User wants maximum rigor | → /certainty $ARGUMENTS |
For binary decisions: → INVOKE: /araw [doing X vs not doing X]
For multi-option with known criteria: → INVOKE: /cmp [options with criteria]
For multi-option where dominance might simplify: → INVOKE: /dom [options] — then /cmp on survivors
For open decisions (options unknown): → INVOKE: /search $ARGUMENTS — discover options → Then INVOKE: /cmp [discovered options]
For high-stakes decisions, also: → INVOKE: /prm [chosen option] — imagine it failed, trace back why → INVOKE: /fla [chosen option] — anticipate failure modes
For decisions where criteria are unclear: → INVOKE: /want $ARGUMENTS — find the underlying goal → Then return to /decide with clarified criteria
| Failure | Signal | Fix |
|---|---|---|
| False binary | "Should I X?" tested as yes/no only | Look for the third option — the reframe, the hybrid, the delay |
| Criteria-free comparison | "X is better" without stating better-for-what | Stop and clarify criteria before proceeding |
| Analysis paralysis | Reversible decision getting deep analysis | Match depth to stakes — reversible decisions need light analysis |
| Insufficient depth | Irreversible decision getting light analysis | Irreversible decisions need /araw + /prm + /fla minimum |
| Foreclosure blindness | Only looking at what options gain, not what they give up | Map foreclosures explicitly for every option |
| Framing lock | All options feel wrong but analysis continues | Step back — wrong options usually mean wrong framing. Reframe what's actually being decided |
| Depth | Scope | Output |
|---|---|---|
| 1x | Quick — state options, criteria, recommendation | Options listed, criteria applied, quick verdict |
| 2x | Standard — /cmp with explicit criteria, foreclosures noted | Full comparison table, foreclosures for each, recommendation with confidence |
| 4x | Thorough — /cmp + /araw on top candidate + assumption check | Comparison + stress test of winner + assumptions surfaced |
| 8x | Deep — /araw each option + /prm + /fla + alternative framing check | Full analysis of all options, failure modes mapped, foreclosures, reframe explored |
Report:
After the decision is made, the user may need:
