Skip to content

Cognitive Operating System Extraction

Cognitive Operating System Extraction is a methodology for systematically deconstructing an individual's decision-making architecture and problem-solving style into a programmable format[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md][001-TODO__nuwa-skill_-Claude_Code_认知框架提取工具.md]. Unlike simple role-playing or persona generation, which relies on mimicking surface-level tone, this process aims to extract the underlying "source code" of a person's mind—their mental models, heuristics, and cognitive frameworks[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md][001-TODO__nuwa-skill-_Claude_Code_认知框架提取工具.md].

The extracted "OS" can then be packaged as an [[AI Skills]] module, allowing users to apply a specific expert's perspective (e.g., Charlie Munger, Elon Musk, or Paul Graham) to analyze real-world problems in a predictable and consistent manner[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md][001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].

The Five-Layer Model

A robust cognitive operating system is composed of five distinct layers^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md]:

  1. Expression DNA: The individual's unique communication style, including tone, rhythm, and diction.
  2. Thinking Process: The mental models and cognitive frameworks used to structure information.
  3. Decision Logic: The heuristics and rules of thumb applied to make judgments quickly.
  4. Anti-Patterns: Values and behaviors that the individual explicitly avoids (what they don't do).
  5. Honesty Boundaries: The defined limits of the system, acknowledging areas where the model cannot make accurate predictions^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].

The Extraction Process

The creation of a cognitive OS follows a structured workflow involving parallel data gathering, rigorous filtering, and validation[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md][001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].

1. Parallel Data Collection

Six distinct "agents" or streams are used to gather information simultaneously to ensure a comprehensive view^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md]: * Publications and books * Podcasts and interviews * Social media activity * Critic perspectives * Decision records * Life timeline

2. Triple-Verification Filtering

Not every opinion or quote constitutes a valid part of the operating system. To be included as a core mental model, a concept must pass three criteria^[001-TODO__nuwa-skill_-Claude_Code_认知框架提取工具.md]: * Cross-Domain Recurrence: The concept must appear in two or more different domains, ensuring it is a fundamental belief rather than a situational comment^[001-TODO__nuwa-skill-Claude_Code_认知框架提取工具.md]. * Predictive Power: The model must be capable of inferring the individual's stance on novel, unseen problems^[001-TODO__nuwa-skill-Claude_Code_认知框架提取工具.md]. * Exclusivity: The perspective must be unique to the individual, distinguishing them from generic "smart" consensus^[001-TODO__nuwa-skill-_Claude_Code_认知框架提取工具.md].

3. Quality Assurance

The extracted system is tested against known answers to verify alignment. It is also tested on questions the individual has never publicly addressed to ensure it exhibits appropriate uncertainty rather than hallucinating an answer^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].

Practical Applications

Once extracted, the cognitive OS serves as a functional tool for decision-making and creativity[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md][001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md]:

  • Product Design: Using [[Steve Jobs]]' or [[Zhang Yiming]]'s frameworks to critique product direction^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].
  • Communication: Applying [[Paul Graham]]'s or [[Richard Feynman]]'s expression DNA to improve writing or reporting clarity^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].
  • Investment: Utilizing [[Charlie Munger]]'s multi-disciplinary思维模型 to analyze risk and opportunity^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].
  • Engineering: Employing [[Elon Musk]]'s first-principles reasoning to optimize system architecture^[001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md].
  • [[Mental Models]]
  • [[Heuristics]]
  • [[Prompt Engineering]]
  • [[AI Agents]]

Sources

  • 001-TODO__nuwa-skill_-_Claude_Code_认知框架提取工具.md