BearPack Projects

Games, autonomous agents, simulation technology, creator tools, and experimental platforms built through BearPack and Arctic Codex.

Featured projects

BearFront.io strategy game artwork
Active DevelopmentGames

BearFront.io

A fast territorial browser strategy game where every border becomes a tactical decision.

BearFront.io is a real-time world-domination game built around territorial expansion, economy management, contested borders, cities, ports, naval warfare, and short strategic matches.

Smooth territorial conquestLand and naval strategyBots and tactical economy
StoneSeed research prototype artwork
ResearchArtificial Intelligence

StoneSeed

An autonomous Minecraft agent designed to reason, survive, learn, and eventually complete the game.

StoneSeed is an experimental game-playing AI that observes Minecraft, selects goals, maintains multi-step intentions, uses executable skills, verifies real world-state changes, remembers outcomes, and responds to emergencies without relying on constant player commands.

Persistent goals and intentionsVerified actions and reward learningCross-game adapter architecture

Active and notable projects

Unified Physics Engine technical artwork
Technical PreviewSimulation & Engines

Unified Physics Engine

A deterministic native simulation runtime built for repeatability, replay, rollback, and inspectable behaviour.

Unified Physics Engine is a native physics and simulation project centred on deterministic stepping, state hashing, snapshots, replay workflows, and a stable integration boundary.

Deterministic simulationSnapshots and replayNative commercial source architecture
BearPack OpenFront AI research artwork
ResearchArtificial Intelligence

BearPack OpenFront AI

A research platform for training strategic agents through imitation learning, reinforcement learning, self-play, and cross-game transfer.

BearPack OpenFront AI uses OpenFront as the first reference environment for a larger general game-intelligence platform. The goal is to build reproducible strategic agents that can learn from demonstrations, compete through self-play, model opponents, plan ahead, and transfer knowledge to...

Imitation learning and PPOStrategic self-playCross-game agent architecture
Web2TXT browser research tool artwork
Working PrototypeCreator Tools

Web2TXT Research Suite

A local-first browser research tool for turning tab-heavy browsing sessions into organized text.

Web2TXT Research Suite captures readable text from eligible tabs in the current browser window and exports the material into a structured local text file for research, archiving, documentation review, and AI prompt preparation.

Multi-tab text exportTitles, URLs, and source contextLocal-first research workflow

Long-Term Projects

Substantial work that should be understood as paused or longer-horizon rather than launch-ready.

ArctiCasters.tv streaming platform artwork
PausedPlatforms

ArctiCasters.tv

A creator-first live-streaming platform designed around community, discovery, monetization, and creator control.

ArctiCasters.tv is a large streaming-platform project that combines live channels, chat, discovery, creator dashboards, moderation, scheduling, analytics, subscriptions, donations, tipping, and an eventual on-site stream builder.

Live streaming and realtime chatCreator dashboards and monetizationDiscovery, moderation, and analytics

Experiments and earlier work

Experimental

All Tabs Open / Bulk Link Opener

Research-family browser utility for previewing, filtering, deduplicating, and opening useful links from large tab sets.

Learn More
Experimental

KodiakLocalAI

Local AI experiment work that is still incubating while flagship systems are prioritized.

Learn More
Working Prototype

Pack Pulse

Playable rhythm experiment with a live route that remains intentionally secondary to flagship projects.

Play Demo
Experimental

BearPackAI Core

Foundational AI orchestration work supporting broader agent systems and tooling research.

Learn More
Experimental

GameObserver

Small observation and monitoring experiment tracked as earlier-stage supporting work.

Learn More
Experimental

AutoLearner

Early learning experiment that remains part of the incubator track.