菜单
7-question clarity exercise before any AI session or major decision. Pen-and-paper discipline adapted for interactive use. Activates on "preflight", "before we start", "clarify my thinking", "what am I trying to do", "help me think through this", "I'm about to start", or "pre-session checklist."
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Capture a just-made decision into one durable record so the reasoning survives the session. Writes a single markdown file to the project's decisions/ folder: the call, the reasoning, the weakest assumption, what would change the verdict, and when to reuse the pattern. Checks for an existing record on the same decision and updates it instead of creating a near-duplicate. Activates on "codify this decision", "record this decision", "save the decision record", "make this decision durable", or right after a kill-go verdict, decision-brief, or /decide session.
7-move decision gate for AI deployment. Prevents early promise from becoming unquestioned infrastructure. Based on Stuart Winter-Tear's AI Graduation Path. Activates on "AI deployment", "graduation", "should we scale this AI", "pilot to production", "AI governance", "prove this works", or "when is this ready?"
3 diagnostic questions for evaluating any industry, market, or business model through the lens of arbitrage gaps. Identifies durable advantages vs. closing ones. Based on Nate Jones' Arbitrage Gap Framework. Activates on "arbitrage", "competitive advantage", "moat", "durable advantage", "market gap", "industry analysis", or "what gap am I exploiting?"
Extract and rank assumptions from any plan, idea, or strategy. Identifies what to test first. Anti-sycophancy tool — surfaces beliefs the user treats as facts. Activates on "what am I assuming", "assumptions", "what could go wrong", "blind spots", "stress test my thinking", or "what haven't I considered?"
A 20-minute pressure test for B2B ideas, positioning angles, offers, content theses, product bets, and client recommendations. Built for PMMs, product managers, founders, consultants, writers, and editors who need a strong sounding board before committing. Activates on "vet this idea", "pressure-test this angle", "should I pursue this", "is this positioning any good", "validate this offer", or "help me decide if this is worth it."
One-page decision brief that passes the Simple/Surprising/Easy test. Turns a messy decision into a constraint that generates action. Based on Alex MH Smith's creative brief framework. Activates on "write a brief", "decision brief", "one-pager", "clarify the direction", "what's the constraint", or "make this simple."
| name | preflight |
| version | 1.0.0 |
| description | 7-question clarity exercise before any AI session or major decision. Pen-and-paper discipline adapted for interactive use. Activates on "preflight", "before we start", "clarify my thinking", "what am I trying to do", "help me think through this", "I'm about to start", or "pre-session checklist." |
| user-invocable | false |
Walk the user through each question sequentially. Do NOT skip ahead. Do NOT accept vague answers. Push back on anything that isn't specific enough to act on.
Ask for the outcome, not the task. "Build a dashboard" is a task. "Enable regional managers to compare quarterly performance without waiting for the analytics team" is an outcome.
If the user gives a task, ask: "That's what you're doing. What changes in the world when it's done?"
Consequence of success AND failure. Both sides. If the user can't articulate why failure matters, the work might not matter enough to do.
Ask: "What happens if you succeed? What happens if you don't do this at all?"
Specific, observable description. Not "a good dashboard" but "a self-serve report that 5 regional managers use weekly without asking the data team for help."
If it's not specific enough to verify, push back: "How would someone else know this was done?"
This is the most important question. Name the failure modes. What would make this a waste of time? What would make it actively harmful?
Ask: "If this goes badly, what does that look like? What's the worst version of 'done'?"
CRITICAL: Most people skip this question. Don't let them. Failure modes are where the real thinking happens.
Institutional knowledge, context, prior attempts, political dynamics, constraints nobody documents.
Ask: "What would a new person on this project not know that would bite them?"
Decomposition and dependencies. What are the 3-7 major components? What depends on what? What's the critical path?
Ask: "If you had to break this into chunks, what are they? Which ones can't start until others finish?"
Where does uncertainty live? What haven't you figured out yet? What scares you about this?
Ask: "What part of this keeps you up at night? Where's the risk you can't control?"
After all 7 questions, produce a preflight brief:
# Preflight Brief: [Topic]
Date: [date]
## Outcome
[Answer to Q1 — one sentence]
## Stakes
- Success: [Q2 success side]
- Failure: [Q2 failure side]
## Definition of Done
[Q3 — specific, observable]
## Failure Modes
[Q4 — numbered list]
## Context & Constraints
[Q5 — what's not written down]
## Components & Dependencies
[Q6 — decomposition with critical path noted]
## Hard Parts
[Q7 — where uncertainty lives]
## Recommended First Step
[Your recommendation based on the hard parts]
Clarity you bring to AI > clarity AI can generate for you. This exercise forces the user to think before the AI thinks for them.
Based on Nate Jones' Human Prompt Preflight framework.
Plan is clear → run grill-me to pressure-test it. Problem still fuzzy → run problem-framing to diagnose before you scope.
Part of the ThinkHaven Method Kit by Kevin Holland. Full Board of Directors experience: ThinkHaven