Storm Research Protocol
Overview
This skill implements the Stanford OVAL STORM (Synthesis of Topic Outlines through
Retrieval and Multi-perspective question generation) framework as a 4-stage prompt protocol.
It forces a research task through five deliberately opposed expert lenses, surfaces the
contradictions and blind spots between them, synthesizes a CEO-grade briefing, and then
peer-reviews its own output for reliability and bias. Result: a PhD-level due-diligence
briefing produced in minutes rather than days.
The skill's value is the questioning protocol, not the topic. Use it whenever a single
question would otherwise yield shallow, consensus-level "correct nonsense."
When To Use
- Deep industry / market / competitive research
- Investment or project due diligence (看多 vs 看空交锋)
- CEO / exec-level decision briefs on an unfamiliar topic
- Interview prep, negotiation prep, or exec-readiness defense
- Rapidly getting up to speed in a new domain (filter marketing noise)
- Any task where the user explicitly asks for multi-angle, rigorous, evidence-ranked analysis
If the user only wants a quick factual lookup, do NOT invoke this skill — it is heavyweight by design.
Workflow
- Capture inputs. Identify the research topic (replace
[YOUR TOPIC]) and, for Stage 3,
the reader's role (replace[YOUR ROLE]). If the user did not specify a role, infer a
sensible default (e.g. "投资决策者" / "业务负责人") and state the assumption. - Run the 4 stages sequentially. Execute Stage 1 → feed its output into Stage 2 → feed
both into Stage 3 → feed the Stage 3 briefing into Stage 4. Each stage below is a verbatim
prompt set; keep the section structure and deliverables exactly as written. - Produce one consolidated deliverable. Present the final output as a single structured
report with four labeled sections (Five-Perspective Scan, Contradiction & Blind-Spot Map,
CEO Briefing, Peer Review). Use A-share color convention (涨红跌绿) only if financial data appears. - 5-minute cadence (optional framing for the user): Stage 1 ≈ 1 min, Stage 2 ≈ 2 min,
Stage 3 ≈ 1 min, Stage 4 ≈ 1 min. The whole pipeline is ~300 seconds.
Stage 1 — 五维多视角扫描协议 (Five-Perspective Scan)
> 任务目标
> I need to research [YOUR TOPIC 替换你的调研主题].
> Simulate 5 different expert perspectives on this topic:
1. THE PRACTITIONER（一线实操者）
• works with this daily.
• What do they know that academics miss?
• What practical realities are usually ignored?
2. THE ACADEMIC（学术学者）
• has studied this for years.
• What does the peer reviewed evidence actually say?
• Where does the evidence contradict popular belief?
3. THE SKEPTIC（死硬怀疑论者）
• thinks the mainstream view is wrong.
• What is the strongest counterargument?
• What evidence do proponents conveniently ignore?
4. THE ECONOMIST（利益与经济学家）
• follows the money.
• Who profits from the current narrative?
• What financial incentives shape the research?
5. THE HISTORIAN（周期历史学者）
• has seen similar patterns before.
• What historical parallels exist?
• What can we learn from how those played out?
> 输出规范 · 每个角色交付 3 样东西
> • Their core position in 2 sentences（两句话核心立场）
> • The strongest evidence supporting their view（支撑其观点的最硬核证据）
> • The one thing they would tell me that no other perspective would（任何其他角色都绝不敢说的一句大实话）
Stage 2 — 矛盾与盲区地图映射 (Contradiction & Blind-Spot Map)
> 任务目标
> Based on the 5 perspectives above, map the contradictions:
1. 抓直接冲突（Direct Conflicts）
Where do two or more perspectives directly contradict each other? List each conflict with the specific claims that clash.
2. 评证据强弱（Evidence Strength）
Which perspective has the strongest evidence? Which has the weakest? Why?
3. 定胜负天平（The Keystone Question）
What is the one question that, if answered, would resolve the biggest contradiction?
4. 提全员共识（Universal Agreement）
What does EVERY perspective agree on?
（注：连死对头都一致承认的事实，大概率就是真金白银的真相）
5. 挖全场盲区（The Blind Spot）
What topic did NONE of the perspectives address?
（注：全场都没提到的盲区，往往藏着整个行业最大的认知红利）
> 两条黄金法则
> 1. 若 5 个对立立场全部认同某事 → 大概率是不可撼动的客观事实。
> 2. 若 5 个专家无人提及某关键维度 → 你抓到了行业最大的认知盲区与价值洼地。
Stage 3 — 总裁级极速决策简报协议 (CEO-Level Briefing)
> 任务目标
> Synthesize everything from the 5 perspectives and the contradiction map into a research briefing:
1. 60 秒极简电梯速览（The One-Paragraph Summary）
Explain this topic as if briefing a CEO who has 60 seconds and needs nuance, not just the headline.
2. 五大关键发现按可靠度严谨排序（The 5 Key Findings）
Most important things I now know, ranked by reliability. For each, note which perspectives support it and which challenge it.
3. 跨视角的非显性隐藏连接（The Hidden Connection）
One non-obvious link between findings that only shows up when you look at all 5 perspectives together.
4. 针对具体岗位的行动建议（The Actionable Insight）
Based on all the evidence, what should someone in [YOUR ROLE 你的岗位角色] actually DO differently? Be specific.
5. 改变行业格局的前沿天问（The Frontier Question）
The one question that, if answered, would change everything about how we understand this topic.
Stage 4 — 严苛同行评审与事实质检 (Peer Review & Fact-Check)
> 任务目标
> Now peer review your own research briefing:
1. 置信度评分打假（Confidence Scores）
Rate each of the 5 key findings on a 1 to 10 scale for reliability. Explain each score.
2. 揪出最脆弱的一环（Weakest Link）
Which claim are you least confident in? What specific info would you need to verify it?
3. 偏见审查防带节奏（Bias Check）
Which perspective might be overrepresented in your synthesis? Did one voice dominate?
4. 遗漏视角补充检查（Missing Perspective）
Is there a 6th angle I should have included that would change the conclusions?
5. 斯坦福教授级严苛终评（Overall Grade）
If a Stanford professor reviewed this briefing, what grade would they give and why? What would they tell me to fix?
High-Value Scenarios (reference for the user)
- 撰写深度研报 — 提前覆盖盲区角度，论点密不透风
- 重大商业决策 — 老手讲实操，怀疑论者挑硬伤，经济学家挑破利益局
- 关键求职面试 — 内行话 + 高维穿透性提问
- 重大项目投资 — 多空交锋、历史周期对比、矛盾地图锁定风险
- 快速掌握新领域 — 过滤营销噪音，直击底层逻辑
- 关键商务谈判 — 摸透对方动机与弱点
- 高管汇报答辩 — 提前闭环所有质疑与反驳
Notes
- Keep the original English prompt phrasing where provided; it is tuned for model compliance.
- This protocol works on any topic — the rigor comes from the five opposed lenses, not from tooling.
- Stage 4 is mandatory. Skipping self-review is the most common failure mode that lets biased
or hallucinated conclusions pass as findings.
