"How to work with me" as a reusable AI skill for Max Lungarella, also known as cybermax and Hulk. Inspired by the manager README tradition, but written to be loadable by AI systems as well as humans.
“The future of AI is not confined to code. It is embodied, relational, and meaningful only insofar as it advances the dignity, health, and flourishing of human life in the universe.”
— Max Lungarella
This repository is a working guide to Max Lungarella's standards, priorities, and communication style.
It is meant to help collaborators, assistants, and AI tools produce work that is sharper, more human, and more useful when Max is the audience, author, subject, or decision-maker.
The core file is skills.md. If you want Claude-compatible packaging, rename that file to SKILL.md before loading it.
Use this repository when you are:
- writing to Max Lungarella
- preparing material that Max will review
- drafting messages to Max
- creating content about Max, cybermax, or Hulk
- aligning the tone of an AI assistant with Max's standards
- packaging a reusable personal skill for deep-tech collaboration
The skill is designed to make Max's working style explicit.
It covers:
- identity, mission, and worldview
- communication and writing standards
- preferred tone and depth
- feedback style
- decision-making patterns and prioritization logic
- anti-patterns in language and reasoning
- collaboration and meeting norms
README.md: human-readable overview and usage guideskills.md: the actual behavioral skill definition
Copy the files into your own repository and adapt the content to your own voice, background, and operating style.
Rename skills.md to SKILL.md and place the folder in your Claude skills directory.
Example:
git clone https://github.com/your-username/cybermax-skill.git ~/.claude/skills/cybermax-skillThen load the skill whenever Max is the audience, author, or subject of the work.
Too much useful context in deep-tech organizations remains tacit.
People often do not know how a founder thinks, what level of rigor they expect, what tone they trust, or which habits signal shallow thinking. AI systems know even less unless those preferences are made explicit.
This repository turns that tacit context into operational infrastructure for better collaboration.
It is not a personal branding exercise. It is a compact tool for reducing noise and increasing signal.
Fork this repository and replace the content with your own background, principles, and standards.
Do not try to sound grand. Try to be legible.
A good personal skill answers questions like:
- What do you care about?
- How do you think?
- How should people write for you?
- What makes you trust a document?
- What counts as weak thinking?
- How should collaborators challenge you?
If those answers are clear, both humans and machines can work better around you.
Apache 2.0. Fork it, adapt it, and make it useful.