DEV Community

Daniel Ioni
Daniel Ioni

Posted on

From GitHub to the Metaverse: Building Verified Profiles and Knowledge with MyZubster and Zorgax

From GitHub to the Metaverse: Building Verified Profiles and Knowledge with MyZubster and Zorgax

Over the last few days, we have been working on a part of MyZubster that is becoming increasingly important to the whole project:

turning a simple user account into a verified digital identity connected to real skills, experiences, projects, evidence, contributors and knowledge.

The goal is not to build another social network where people write a short bio and stop there.

The idea behind MyZubster is broader.

A profile should be able to explain:

  • who you are;
  • what you can actually do;
  • what you are learning;
  • what projects you have worked on;
  • what evidence exists;
  • who you have collaborated with;
  • which parts of your experience are verified;
  • and which information you explicitly want to make public.

At the same time, MyZubster should avoid inventing skills or turning weak signals into professional qualifications.

That is where Zorgax enters the picture.


GitHub as a verified identity layer

One of the systems we are developing is the connection between MyZubster and GitHub.

When a user authenticates through GitHub, MyZubster can associate the account with a verified GitHub identity.

That is very different from simply allowing someone to write:

"My GitHub username is this."

Instead, MyZubster can know that the GitHub identity is actually connected to the authenticated account.

From there, Zorgax can use publicly available information such as:

  • GitHub profile information;
  • repositories;
  • profile README;
  • repository descriptions;
  • programming languages;
  • public project activity.

But an important principle is:

data is not automatically a skill.

Having a JavaScript repository does not automatically mean someone is an advanced JavaScript developer.

Having cybersecurity repositories does not automatically make someone a professional security engineer.

The system should distinguish between:

  • evidence;
  • declared experience;
  • verified activity;
  • inferred context;
  • and actual professional claims.

This distinction is very important for us.


Yassen: from GitHub authentication to the MyZubster Metaverse

We have already seen this workflow become real with Yassen.

Yassen configured his GitHub profile and connected it to MyZubster.

Through that process, he entered the MyZubster Metaverse with a complete authenticated identity.

This is important because the Metaverse character is not just an anonymous avatar disconnected from everything else.

The identity can be linked back to the authenticated MyZubster account and its verified GitHub profile.

So the flow becomes something like:


text
GitHub identity
        ↓
MyZubster authentication
        ↓
verified account
        ↓
Metaverse character
        ↓
projects, contributors and evidence
Yassen also added danieldirimini-myzubster as a contributor.
That creates another important connection.
Instead of thinking about profiles as isolated pages, we can start thinking about a network of people who have actually collaborated.
A contribution can become part of a larger graph:
Person
  ↓
Project
  ↓
Contribution
  ↓
Evidence
  ↓
Other contributors
This is one of the directions we want MyZubster to develop further.
Profiles should describe real people, not AI-generated identities
While building the profile onboarding system, we also found an interesting problem.
Zorgax is an AI assistant inside MyZubster.
During early profile onboarding tests, a question such as:
"What practical activities can you do?"

could sometimes be interpreted by the assistant as a question directed at Zorgax itself.
That produced answers about what the AI could do instead of interviewing the user.
We corrected this behavior.
The onboarding system now explicitly tells Zorgax:
- it is interviewing the user;
- the questions are about the user's professional profile;
- it must not answer questionnaire questions as itself;
- it must not attribute Zorgax capabilities to the user;
- it should ask one question at a time;
- it should wait for the user's answer.
This sounds like a small change, but it represents a fundamental rule.
AI must help structure a person's identity, not create one for them.
Building my own professional profile inside MyZubster
We are now using the same system to structure my own profile.
Instead of creating a generic biography, we went through my experiences one by one.
We separated:
- skills I can use operationally;
- long-term practical experience;
- self-taught knowledge;
- things I am currently learning;
- available evidence;
- interests;
- professional goals;
- collaboration availability.
We also deliberately avoided presenting self-taught experience as formal certification.
For example, my profile includes areas such as:
MyZubster
I directly manage the development and evolution of MyZubster.
My work includes:
- project coordination;
- testing;
- identifying problems;
- following deployments;
- documentation;
- coordinating contributors;
- using GitHub, Linear and Vercel;
- using AI tools such as ChatGPT and Zorgax;
- working with the VPS and terminal when necessary.
At the same time, I do not describe myself as an advanced software developer if that level of expertise has not yet been reached.
That distinction matters.
Cybersecurity and Kali Linux
I have spent more than ten years studying and experimenting with Kali Linux and cybersecurity in a self-taught way.
The public professional profile only represents legitimate skills such as:
- Linux;
- terminal usage;
- networking;
- scanning;
- configuration;
- security testing in authorized environments or laboratories.
The profile system is designed to exclude unrelated or sensitive information that should not become part of a public professional identity.
Calisthenics
I practiced calisthenics for approximately four years.
My practical experience includes movements such as:
- human flag;
- L-sit;
- V-sit;
- muscle-up;
- freestyle;
- back lever;
- front lever.
There are video records available as evidence.
Again, the system does not convert this into invented certifications or coaching qualifications.
It simply records the experience that actually exists.
Thai boxing and MMA
My profile also includes:
- around four years of Thai boxing;
- around two years of MMA;
- training and sparring with other athletes.
This is recorded as practical experience, not as professional coaching or competitive certification.
Event organization and free parties
Another significant part of my experience comes from events.
I have participated in free party culture for almost twenty years and personally organized more than ten events.
This involved practical responsibilities such as:
- finding locations;
- coordinating people;
- DJs and music;
- sound systems;
- bar organization;
- shifts;
- communication;
- security;
- transport;
- technical setup;
- visual setup.
I have also worked directly with:
- lasers;
- LaserCube;
- projectors;
- screens;
- visual effects.
I am also learning video mapping.
For audio systems, I have practical basic knowledge of:
- connections;
- mixers;
- speakers;
- levels;
- simple troubleshooting.
The profile explicitly avoids describing this as professional audio engineering expertise.
Barista experience
I also have around four years of experience working as a barista.
That includes:
- coffee preparation;
- café work;
- preparation of alcoholic drinks.
Kefir as practical knowledge
Another important area for me is kefir.
I have approximately seven years of practical experience with:
- preparation;
- maintaining kefir grains;
- fermentation;
- managing fermentation time;
- sharing grains;
- teaching other people how to prepare and maintain kefir.
This is particularly interesting for MyZubster because kefir is not only a profile skill.
It can also become part of the knowledge system.
A practical skill can generate:
experience
   ↓
evidence
   ↓
structured knowledge
   ↓
verification
   ↓
knowledge available to others
That is one of the core ideas behind MyZubster Knowledge.
Permaculture, gardening and botany
I have also studied permaculture independently through books and practical experimentation.
My experience includes:
- gardening;
- plants;
- botany;
- practical cultivation.
These activities were also part of my volunteer experience.
Volunteer experience
I spent around two years volunteering with the Comunità Papa Giovanni XXIII.
My practical responsibilities included:
- helping manage the facility;
- organizing work;
- supporting young people;
- coordinating daily activities;
- participating in and coordinating group meetings.
The professional profile deliberately describes this as practical organizational and support experience.
It does not assign medical, psychological or therapeutic professional qualifications that I do not hold.
That is another important principle of the system:
experience should be represented accurately without inflating it into credentials.
Psychology as self-directed study
I have also spent approximately seven years studying psychology independently through books.
This includes interests such as:
- Freud;
- dreams;
- psychological dynamics;
- fragility and personal development.
Again, MyZubster represents this as:
self-directed study
and not as a psychologist or therapist qualification.
The structured professional profile
To support all of this, we added a structured professional profile model to MyZubster.
The profile can contain:
headline
summary
skills
experience level
evidence
practical experiences
interests
professional goals
collaboration availability
approval status
visibility
version
timestamps
One example of a skill can look conceptually like this:
{
  "name": "Kali Linux / Cybersecurity",
  "experience": "10+ years of self-directed study and practice",
  "level": "Self-taught practical experience",
  "evidence": [
    "Personal practice"
  ],
  "notes": "Only legitimate and authorized security activities are represented"
}
Explicit approval before saving or publishing
One of the most important rules we implemented is that Zorgax cannot simply decide to publish a professional profile.
The API requires explicit approval.
A profile update must include something equivalent to:
{
  "approved": true,
  "visibility": "private"
}
or:
{
  "approved": true,
  "visibility": "public"
}
Without explicit approval, the server refuses to register the professional profile.
This means the user remains in control.
The onboarding interface now presents two distinct actions:
Approve and save privately
or:
Approve and publish on MyZubster
The AI can help organize information.
The human decides whether it becomes part of the profile.
Private and public identity are not the same thing
This distinction is very important.
Not everything a person knows, has experienced or has discussed with an AI belongs on a public profile.
For that reason, MyZubster is being designed around different states.
Conceptually:
Private information
       ↓
Profile candidate
       ↓
User review
       ↓
Explicit approval
       ↓
Private profile
       OR
Public profile
This model will also be important for the Knowledge system.
From profiles to knowledge
The next layer is MyZubster Knowledge.
A user's experience should not automatically become universal knowledge.
For example:
"I have made kefir for seven years"

is a personal experience.
It can become evidence for a contribution.
But it is not automatically the same as:
"This is the scientifically correct way to make kefir."

Those are different claims.
So we use another principle:
message != knowledge
candidate != verified
verified != public
A contribution can start as a candidate.
Then it can be connected to:
- evidence;
- sources;
- people;
- projects;
- GitHub activity;
- practical demonstrations;
- reviewers.
Only after verification should it become part of the public Knowledge layer.
Zorgax as the connection layer
Zorgax is becoming the system that connects these pieces.
It is important to clarify that Zorgax is not simply "a language model trained from scratch."
Zorgax is the AI and orchestration layer inside MyZubster.
It can combine:
- language models;
- verified knowledge;
- evidence;
- workflows;
- user profiles;
- GitHub;
- MyZubster projects;
- contributors;
- human approvals.
So the architecture starts looking more like:
                MyZubster
                    |
        +-----------+-----------+
        |                       |
      Users                  Knowledge
        |                       |
    Profiles                 Evidence
        |                       |
      GitHub                  Sources
        |                       |
 Contributions ------------ Verification
        |                       |
        +--------- Zorgax ------+
                    |
                Metaverse
The Metaverse becomes more than avatars
This is where the Yassen example becomes especially interesting.
If a Metaverse user can be connected to:
- an authenticated MyZubster account;
- a verified GitHub profile;
- real repositories;
- contributions;
- other contributors;
- professional knowledge;
- evidence;
then the Metaverse stops being just a virtual world populated by anonymous avatars.
A character can become a representation of a real digital identity.
Imagine entering the MyZubster Metaverse and meeting someone.
Instead of only seeing:
Yassen
the system could eventually allow you to understand:
Yassen
├── Verified MyZubster account
├── Verified GitHub identity
├── Projects
├── Contributions
├── Collaborators
├── Skills
├── Knowledge contributions
└── Evidence
Of course, only information the user has chosen to make public should appear.
Contributors as part of the identity graph
When Yassen added danieldirimini-myzubster as a contributor, it demonstrated another direction for the ecosystem.
A project is not only code.
It is a network of people.
That network can eventually connect:
Developer
   ↓
Repository
   ↓
Contribution
   ↓
Contributor
   ↓
Evidence
   ↓
Reputation
   ↓
Knowledge
This can become much more meaningful than follower counts.
Instead of asking:
"How popular is this person?"

we can ask:
"What have they actually contributed?"

Why we are doing this
I want MyZubster to become my main professional project.
But at the same time, I am using the project itself as a way to continuously develop my skills.
That means MyZubster is also becoming a record of the learning process.
Instead of pretending to already know everything, the profile can distinguish between:
I know this
I have practical experience with this
I am currently learning this
I have evidence for this
I want to improve this
For me, this is much more honest and useful than a traditional CV.
The bigger idea
What we are building can eventually connect four important concepts:
Identity
Who is this person?
Evidence
What proof exists?
Contribution
What has this person actually done?
Knowledge
What can the community verify and reuse?
And Zorgax can help connect them without replacing human judgment.
The long-term model looks something like this:
IDENTITY
   ↓
EXPERIENCE
   ↓
EVIDENCE
   ↓
CONTRIBUTION
   ↓
VERIFICATION
   ↓
KNOWLEDGE
   ↓
COLLABORATION
That is the direction we are exploring with MyZubster.
Not just a marketplace.
Not just a Metaverse.
Not just an AI assistant.
Not just GitHub integration.
But an ecosystem where people, projects, real-world experience, evidence and knowledge can become connected.
And, most importantly, where the person remains in control of what becomes public.
What we have working now
At the current stage we have already implemented important pieces of this architecture:
- GitHub-based verified identities;
- authenticated MyZubster accounts;
- GitHub-linked Metaverse characters;
- structured professional profiles;
- skills and experience with evidence;
- explicit user approval before profile persistence;
- private and public profile visibility;
- Zorgax-guided profile onboarding;
- protection against AI inventing user capabilities;
- a public Knowledge Explorer;
- a verification-oriented Knowledge layer;
- contributor relationships;
- Vercel deployment infrastructure.
The professional profile implementation has also been tested with automated regression tests before deployment.
What we are doing next
The next step is not simply to add more fields.
The real challenge is connecting the graph.
We want to increasingly connect:
MyZubster profile
       ↓
GitHub
       ↓
projects
       ↓
contributors
       ↓
evidence
       ↓
Knowledge
       ↓
Zorgax
       ↓
Metaverse identity
My own profile is currently being used as one of the real examples for building this system.
The goal is to make the profile itself useful to the ecosystem:
not just a biography,
but a structured representation of what I can do, what I have done, what I am learning, what evidence exists and how I can collaborate with others.
That is the direction we are taking MyZubster.
Enter fullscreen mode Exit fullscreen mode

Top comments (0)