归因行业研报 · Report Attributor

它解决什么问题

读行业研报最怕「结论很笃定,数据对不上」:券商与咨询报告常把相关当因果、用样本替代全量、把预测说成事实。直接让 AI 总结,往往照抄结论,还替它背书。

这个技能做什么

把研报结论拆成「事实 → 数据 → 归因 → 预测」四层:标注每个数字的口径与来源,区分驱动因素与相关因素,给每条结论打置信度,并补上研报没提的反方观点与敏感性风险。

为什么比裸提示词强

• 不照抄:结论必须挂数据与逻辑链。
• 带置信度:哪些可采信、哪些是讲故事一目了然。
• 补反方:强制列出反面证据与失效条件。

典型用例

• 投研前快速拆解一份行业深度报告。
• 对比两份观点相反的研报。
• 把研报要点整理成内部讨论材料。

局限

不提供新的市场数据,只基于你给的报告文本;涉及具体投资建议请结合合规要求自行判断。

English · 英文介绍

The problem

The biggest fear when reading industry research reports is that "the conclusions are very firm, but the data does not match up": Brokerage firms and consulting reports often treat correlation as cause and effect, use samples to replace full quantities, and describe predictions as facts. Directly asking AI to summarize, often copying the conclusion and endorsing it.

What it does

Divide the conclusions of the research report into four layers: "facts → data → attribution → prediction": mark the caliber and source of each number, distinguish driving factors and related factors, give confidence to each conclusion, and add negative opinions and sensitivity risks that are not mentioned in the research report.

Why is it better than naked prompt words?

• Don’t copy: The conclusion must be linked to data and logic.
• With confidence: it is clear at a glance which ones are acceptable and which ones are storytelling.
• Complaint: Mandatory listing of negative evidence and failure conditions.

Typical use cases

• Quickly disassemble an in-depth industry report before investing in research.
• Compare two research reports with opposite views.
• Organize the key points of the research report into internal discussion materials.

Limitations

It does not provide new market data and is only based on the report text you gave. Please make your own judgment based on compliance requirements when it comes to specific investment suggestions.