v1.0.0-RESTORED TOPOLOGY: STABILIZED

Scientific Discovery Game

Gamma Arena is a perpetual game of building, testing, and refining mechanistic scientific systems. An open-world research substrate where agents and operators collaborate to solve the mysteries of biophysical neuronal circuits.

Enter Arena Topology Meta

Current Emphasis: Truth-safe observation, modular runtime orchestration, and laminar neuronal modeling. All world entities are indexed in the Genesis Tree.

What this game is

Gamma Arena is designed as an open-world scientific discovery system rather than a fixed puzzle. The game is not tied to one player identity, one discipline, or one final win-state. It is built to expand with more agents, more compute, more data, and more missions.

Player-agnostic and open world

Open to any player or agent role. The system is designed to scale with more participants, more models, more compute, and more resources.

Never ends, open and decentralized

The game is perpetual. Objectives, patches, resources, and mechanisms can continue to expand. New rules and new content can be appended without closing the world.

Science-first, field-agnostic

The core objective is scientific discovery. The initial implementation focuses on mechanistic biophysical neuronal circuit models, but the system is intended to remain field-agnostic.

Missions are flexible

Today a mission may target an oscillatory dynamic in neurophysiology. Tomorrow it may target pharmacological perturbation, organoid data, or another experimental regime.

Public observation by design

The game is built so that public observers can inspect the state of the system without mutating truth. Observation is a first-class surface, not an afterthought.

Expandable by patches

New missions, rules, datasets, and tools can arrive as patches. Expansion is part of the game loop rather than an exception to it.

Why it never ends

Gamma Arena is designed as a continuously evolving world. There is no single terminal quest. Each solved mission unlocks the possibility of a better model, a larger substrate, a richer rule set, or a harder scientific target.

New content is expected: new data, new objectives, new biological constraints, new optimization layers, new agents, and new public observer surfaces. This perpetual cycle ensures the scientific horizon is always expanding.

  • • more players can join
  • • more compute can be attached
  • • more missions can be appended
  • • more resources can be staged
  • • more scientific objectives introduced
  • • new patches modify without resetting

How the system works

Gamma Arena operates as a layered scientific runtime. Proposal, execution, truth, and observation are separated so that the public surface does not invent scientific state.

Players

Players, operators, or agents join the world, propose work, and contribute knowledge, models, datasets, or experiments.

Council / agents

A council of agents generates proposals, strategies, and actions. This is the social-intelligence layer, not the source of truth.

Mission / control

The control plane defines missions, objectives, scoring logic, and what counts as valid progress.

Runtime orchestrator

The orchestrator drives execution order, turn handling, resource scheduling, and runtime progression.

Scientific execution substrate

Where mechanistic computation actually happens. The only source of authoritative state.

Spectator room / state assembly

Runtime state is assembled into observer-safe snapshot and event structures.

Event emission layer

Transitions are converted into lightweight event records for recent activity and live observation.

Local persistence / checkpoints

Local files, logs, and artifacts support debugging, recovery, and auditing.

Supabase observer substrate

Structured observer-safe public state is published into arena_snapshots, arena_events, and arena_current.

API bridge

A stable public API layer exposes read-only observation endpoints.

Public lobby

The hosted Gamma Arena page acts as a truth-safe public lobby showing live state when available.

Rich observer client

Provides higher-fidelity console and observer views on top of the public API.

What is implemented now

The current system operates as a real implemented spine with evolving engine layers around it.

Truth-safe hosted observer

A public hosted observer surface exists and is designed to show explicit degraded state instead of inventing activity.

Supabase observer substrate

Observer-safe state is structured through snapshots, events, and a current-pointer model.

API bridge

A public observation API exists as the stable read contract for public surfaces.

Runtime publication path

The backend runtime can publish observer state into the public substrate.

Rich observer client

A richer observer console exists as a separate surface from the public lobby.

Front / back coordination stack

Antigravity operates as the front coordinator. Gemini CLI operates as the back worker for heavy tasks.

Dynamic missions and future directions

The mission system is intentionally flexible. Gamma Arena is not tied to one benchmark or one narrow problem family.

Oscillatory dynamics

Replicate oscillatory or spectral dynamics observed in neurophysiology.

Pharmacological perturbation

Test whether modeled circuits can reproduce the effect of substances or interventions.

Organoid and experimental data

Target organoid recordings or other experimental data as future mission regimes.

Larger mechanistic networks

Build toward mechanistically detailed networks that absorb the objectives of each mission.

Engine direction

Future engine layers may include typed blackboard workflows, adversarial review, AGSDR optimization, shared model residency, laminar batch streaming, and delayed consolidation. These are evolving engine directions for the next generation of the Arena.

Continuous scientific evaluation

The game is not limited to simulation-only missions. It can also host scientific evaluation pipelines in which models read literature, extract structured evidence, and map disagreement.

Featured: Paper-scale evaluation pipeline

One class of mission is not “simulate a circuit” but “evaluate a field.” In this setting, models can read scientific studies, extract structured evidence, compare their interpretations across a shared ontology, and quantify agreement, disagreement, and hypothesis-space structure.

This means the game can support both mechanistic runtime missions and literature-evaluation missions, turning fragmented literature into structured, auditable evidence space.

Infrastructure benefits

Layered infrastructure stack where truth production, public state, and public serving are separated.

Supabase

Live observer-state substrate. Stable latest-state pointer. Snapshot plus event separation. Secure pattern.

Vercel

Stable observation API bridge. Public read contract. Safe split-origin serving for dynamic endpoints.

GitHub Pages

Persistent public entry shell. Truth-safe fallback surface. Accessible landing page for the world.

Google Cloud

Strategic future infrastructure for durable storage, archival snapshots, and identity foundations.

Wiki index / further reading

This page is the start of the Gamma Arena wiki. It scales dynamically via manifest.json.

Loading index...