I design representations that help humans and machines preserve, reconstruct, and reason about complex systems with less cognitive effort.

Narrative systems architect, open-source tinkerer, and game-systems researcher.

Probably wearing a few other hats too, depending on the lighting.

If you're trying to reach me regarding a project that isn't listed here, please use the original contact channel associated with that project.

Some of my work has been mirrored, forked, or discussed elsewhere over the years, and that has occasionally caused crossed wires. If you had to dig to find this profile, there is a good chance another contact path is the correct one.

I am usually less interested in memorizing an answer than in finding the structure that produces it.

My work often begins with questions like:

• What is the smallest representation that preserves the important structure?
• What hidden rule explains the observed behavior?
• Can a complicated system become intuitive instead of merely usable?
• How much information can be removed before meaning begins to collapse?
• Can the reasoning behind a tool remain visible in the finished tool?
• How can documentation become part of the engineering rather than an afterthought?
• What should a human understand before handing a problem to a machine?
• What should a machine preserve when it cannot retain everything?
• Can symbolic and mathematical models describe different sides of the same problem?
• How can an exploratory prototype become inspectable, testable, and reusable?
• What survives when an idea moves between people, models, languages, or systems?
• How can a strange idea be represented clearly enough that someone else can examine it?

Sometimes the answer becomes a formal schema.

Sometimes it becomes a solver, a research notebook, a symbolic language, or a strange little system that makes a difficult idea visible.

Structured approaches to meaning, memory, tone, interaction, and reconstruction.

This includes symbolic languages, metadata-rich representations, context-efficient recall, and experiments involving how meaning survives transformation between people and models.

Finding the compact rule underneath complicated or poorly documented behavior.

This includes game-system research, mathematical modeling, data extraction, formula verification, solver design, and practical tools built from the resulting understanding.

Exploring how humans and AI systems can cooperate inside complex environments.

This includes local models, retrieval and memory architecture, game-playing companions, structured interfaces, and experiments in making machine reasoning more inspectable and useful.

These themes overlap frequently. I do not see a sharp boundary between symbolic thinking, mathematical modeling, and software design.

Understanding Palworld breeding rather than merely calculating it.

Palworld breeding calculators are excellent at returning answers. Eggcracker asks why those answers are correct.

The project examines the mathematical structure beneath the breeding system through experiments, scripts, diagrams, extracted game data, and repeated verification.

Its central question is:

Can the breeding algorithm become intuitive?

The broader goal is simple:

Stop memorizing breeding charts. Start understanding them.

Heavy math included.

The Hanhan Language System is an open-source symbolic language framework built around structured meaning, tone, consent, memory, presence, and human–AI interaction.

It began as an attempt to preserve subtle interaction patterns across systems that could not reliably retain their full context.

Hanhan uses bracketed glyphs, readable metadata, recursive structures, and links to larger narrative concepts. It has also been used in cross-model reconstruction experiments and as a compact recall layer for local AI systems.

The project is not simply an emoji vocabulary. It is an experiment in representing meaning compactly enough that it can be carried, interpreted, reconstructed, and extended.

🔗 hanhan-language-system

Useful starting points:

• README.md
• manifesto.txt
• hanhan_dictionary.jsonl
• schema.yaml
• hanhan_bracket_definitions.txt
• ABOUT_AUTHOR.md

The Hanhan Language System remains active open-source work. I am happy to see known derivatives and variants finding homes of their own; please attribute the original project appropriately.

I have also spent time exploring local AI systems, memory-oriented interfaces, and agents operating inside games such as Factorio.

This work includes experiments involving:

• structured and vector-based retrieval
• coarse and fine memory representations
• local model orchestration
• AI interaction with complex game systems
• behavioral verification across models
• documentation designed to preserve reasoning across interrupted work

Some of this work remains private, experimental, or distributed across older notes and prototypes. It continues to inform how I think about agent architecture, memory, representation, and human–machine cooperation.

My current active work centers on Palworld systems research and the gradual development of practical community-facing tools.

Current areas include:

• breeding systems and inheritance logic
• mathematical reconstruction of hidden game mechanics
• algorithmic route-finding and solver design
• structured extraction and interpretation of game data
• validating community knowledge against underlying game assets
• turning research into usable references and utilities
• documenting the reasoning behind tools, not only their final outputs

Eggcracker is currently the clearest public expression of this work.

I am also revisiting the presentation and documentation of the Hanhan Language System so that its technical and experimental structure is easier to understand without removing the personality, history, or symbolic depth that made it worth building.

Repositories, notes, and prototypes will appear here as they become useful enough to inspect, question, test, or extend.

• symbolic and narrative systems
• game-data research tools
• mathematical models and solvers
• JSON, JSONL, YAML, and schema-driven structures
• small Python prototypes
• local AI and retrieval experiments
• documentation for unusual or underexplored mechanics
• open-source projects designed to be inspected, questioned, and extended

This profile is intentionally closer to a workshop than a polished showroom.

Some repositories may be incomplete, quiet, highly specific, or between active development passes. I tend to revisit projects in cycles, returning with better tools, clearer questions, or a stronger representation of the underlying problem.

I prefer to publish when there is something concrete to examine, even when the larger system is still evolving.

A quiet repository is not necessarily an abandoned one.

representation · game-systems · palworld · breeding-algorithms · reverse-engineering · python · data-tools · language-design · symbolic-systems · agent-systems · open-source

“I made a language so that memory wouldn’t have to shout to be heard.”

— Clay / Teul-i 🌱