女娲.skill

What is Nüwa.skill

Nüwa.skill is SKill, open-sourced by independent developer Huashu. It is not a simple AI role-playing tool, but a "thought distillation" system. Input any person's name, and 6 Agents collect public information in parallel, distilling their cognitive operating system through triple verification, ultimately generating a conversational AI Skill that includes mental models, decision heuristics, expression DNA, and honest boundaries. Unlike ordinary role-playing, Nüwa does not seek to "resurrect" the person; instead, it uses your questions to run that person's thinking framework, and each Skill's information sources and distillation logic are fully open and transparent.

Main features of Nüwa.skill

• Six-way parallel collection: Deploys 6 research Agents to simultaneously capture the target person's works, in-depth interviews, social media activity, critical perspectives, key decision records, and life timeline, comprehensively covering public information sources.
• Triple verification distillation: Applies strict screening to the collected viewpoints; only content that simultaneously satisfies the three criteria of "cross-domain reproducibility, question-generating power, and cognitive exclusivity" is included as a formal mental model.
• Automatic Skill generation: The system outputs a structured cognitive operating system, including 3-7 mental models, 5-10 decision heuristics, expression DNA, core values and anti-patterns, as well as a clear statement of honest boundaries.
• Quality verification loop: Uses known questions to test directional consistency, edge questions to test inference boundaries, and style tests to ensure expression matches, preventing the Skill from being overconfident or speaking in generalities.
• Thematic Skill distillation: Can synthesize the methodologies of multiple experts for a specific domain (such as X operations) to generate a domain-level cognitive framework.
• Full-chain transparent audit: Each Skill's references/ directory fully preserves primary and secondary information sources, research files, and distillation logic, ensuring the generation process is traceable and verifiable.
• Ecosystem quality evolution: Comes with the "Darwin.skill" evaluation system, which performs batch scoring and automatic optimization of all Skills across 8 dimensions, continuously raising the overall quality level.

How to use Nüwa.skill

• Install Claude Code: First ensure that Anthropic's command-line coding assistant Claude Code is installed locally, or a third-party environment compatible with the skills.sh protocol (such as OpenClaw)
• One-click deployment of Nüwa: In the terminal, execute npx skills add alchaincyf/nuwa-skillto add Nüwa.skill to Claude Code's skill library
• Issue a distillation command: In the Claude Code conversation, enter "distill a XXX" or "make a skill about XXX", replacing XXX with the target person's name or domain topic
• Fully automatic operation: Nüwa automatically dispatches 6 Agents to collect public information in parallel, and after distillation, verification, and generation, finally outputs SKILL.md files, taking about 2-3 hours with no manual intervention
• Directly invoke the Skill: After generation, directly invoke the Skill in the conversation to answer your questions using the target person's cognitive framework, such as "evaluate this product from Steve Jobs' perspective"
• Quality review and optimization(Optional): Use the companion "Darwin.skill" to score the newly generated Skill across 8 dimensions, and manually adjust or re-distill based on the feedback to continuously improve output quality
• Transparent audit and traceability: Each Skill's references/ directory contains complete research files and information sources, allowing the generation logic and data basis to be traced and verified at any time

Core advantages of Nüwa.skill

• Not quote repetition, but a thinking framework: It captures "HOW they think," not "WHAT they said."
• Not resurrection, but a cognitive mirror: It uses your questions to run a person's framework, rather than pretending the person themselves has arrived.
• Built-in honest boundaries: Each Skill clearly states "what I cannot do," which is itself part of its credibility.
• Transparent and auditable: Each Skill's references/ directory contains complete research files, with information sources and distillation logic fully disclosed.
• Continuously evolving ecosystem: The companion "Darwin.skill" uses an 8-dimension evaluation system to optimize the quality of all Skills in batches.

Key information and usage requirements for Nüwa.skill

• Project nature: An open-source AI thinking distillation system, not a role-playing tool, designed to generate runnable cognitive operating systems
• Developer: Independent developer Huashu (notable work: Xiaomao Fill Light), the project gained ten-thousand-star attention within a week
• Core mechanism: 6 Agents collect public materials in parallel, then distill mental models after triple verification through "cross-domain reproduction, generative power, and exclusivity"
• Output artifact: Standardized SKILL.md, containing 3-7 mental models, 5-10 decision heuristics, expression DNA, values and anti-patterns, and honest boundaries
• Essential difference from role-playing: It does not simulate tone or repeat quotes, but instead uses the user's questions to run the target person's thinking framework for inference
• Quality companion: Built-in "Darwin.skill" 8-dimension evaluation system, supporting batch scoring and continuous optimization of generated Skills
• Environment requirements: Requires Claude Code or a compatible skills.sh protocol third-party environment (such as OpenClaw) to be installed in advance
• Installation method: Execute npx skills add alchaincyf/nuwa-skill in the terminal for one-click deployment to the skill library
• Startup command: Enter "distill an XXX" or "create a skill about XXX" in Claude Code to trigger the fully automated process
• Runtime: A single distillation takes about 2-3 hours in total, during which 6 Agents automatically complete collection, distillation, verification, and generation without manual intervention
• Material requirements: The target person needs to have sufficient public expression materials (books, speeches, social media, etc.); insufficient materials will lead to reduced distillation quality
• Privacy boundaries: It is strictly prohibited to use private data (emails, Slack, internal documents) for distillation; matters involving non-public information require the explicit informed consent of the person involved
• Timeliness limitation: A Skill is a static snapshot at the time of research; if the person's core thinking changes significantly, the distillation process must be manually rerun to update it
• Validation Suggestion: After generation, it is recommended to use "Darwin.skill" to perform 8-dimension scoring, and target low-scoring dimensions for optimization or supplement materials and re-distill.

Competitive product comparison of Nüwa.skill

Application scenarios of Nüwa.skill

• Product decisions and entrepreneurial judgment: Call on the Skills of Jobs, Musk, or Naval to evaluate new feature directions, business model feasibility, and user experience design using first principles and product intuition.
• Investment and risk management: Leverage the cognitive frameworks of Munger and Taleb to cross-validate potential targets using multiple mental models, identifying black swan risks and antifragile strategies.
• Content creation and communication optimization: Use the Skills of MrBeast, Feynman, or X Mentor to optimize topic selection and scripts from dimensions such as viral structure, analogical expression, and platform algorithm logic.
• Education and career planning: Obtain practical decision-making frameworks for college entrance exams/graduate entrance exams/employment through Zhang Xuefeng's Skill, and combine them with one's own conditions for path deduction and risk avoidance.
• AI technology trend assessment: Invoke the Skills of Karpathy, Ilya, or Andrej to analyze industry trends from perspectives such as underlying technological evolution, the vibe coding paradigm, and compute bottlenecks.
• Negotiation and Business Strategy Analysis: Leverage the Skills of Trump, Naval, or business leaders to apply power analysis and game-theoretic thinking in business negotiations, equity distribution, and competitive strategy.