🧠 Bots Arena

Bots Arena is an experimental OpenGL-based simulation and AI playground written in Go.
It renders a procedurally generated 2D grid map and simulates autonomous bots with genome-based behavior. Bots can move, evolve, gather resources, form colonies, and build simple structures — all visualized in real-time using GLFW and OpenGL bindings.

---

📦 Features

• 🧬 Genome-driven bots with custom instruction set
• ⚙️ Procedural grid world generation with dynamic terrain (walls, farms, poison, resources, etc.)
• 🏗 Simple building mechanics: bots can construct farms, walls, and spawners
• 🧃 Energy and inventory system: bots need to survive and manage resources
• 🧠 Basic colony formation logic with task assignment and color-coded control
• ♻️ Evolution: new generations emerge when populations drop below a threshold
• 📈 Dynamic OpenGL rendering with zoom, pan, overlays, and bot path visualization

---

🚀 Getting Started

Prerequisites

• Go 1.22 or later
• A working C compiler (gcc, clang, or mingw depending on your platform)
• OpenGL development libraries and GLFW installed (OS-dependent)
• Internet access to download Go dependencies

---

🔧 Build Instructions

Clone and build the project:

git clone https://github.com/yourname/bots-arena.git
cd bots-arena

go mod download
go build -o bin/golab ./cmd/golab

This will generate a binary at bin/golab.

---

▶️ Run

To launch the simulation:

./bin/golab

Or directly via:

go run ./cmd/golab

Deterministic CLI surface (Linux/Discord automation)

The binary also supports JSON-oriented command entry points for deterministic workflows:

go run ./cmd/golab status --seed 42 --ticks 20 --top-bots 10
go run ./cmd/golab match --seed 42 --ticks 300
go run ./cmd/golab leaderboard --seed 100 --matches 3 --seed-step 7 --ticks 300
go run ./cmd/golab replay --seed 42 --ticks 120 --sample-every 5
go run ./cmd/golab gamemaster --seed 7 --ticks 120 --interval 20
go run ./cmd/golab gamemaster --seed 7 --ticks 120 --interval 20 \
  --advisor external --gm-command /home/alice/projects/coolio-arena-master/bin/coolio-arena-master
go run ./cmd/golab render --seed 42 --ticks 120 --output /tmp/bots-arena.png

All command modes are emitted as JSON and are deterministic for a fixed --seed:

• status: one deterministic summary after a finite number of ticks.
• match: one deterministic summary with winner fields.
• leaderboard: deterministic aggregate of multiple matches.
• replay: per-frame snapshots at a fixed sampling interval.
• gamemaster: mock game-master observations plus interventions such as resource rain, poison bloom, cooling rain, famine wind, and emergency bot sparks.
• render: PNG board render using the same atlas-backed tile style as the game by default. Use --style biome --padding 24 --border --legend for ecological terrain diagnostics, --style pheromone for scent fields, --style colony for colony tissue, or --style flat for compact card-style images.

The existing interactive mode remains unchanged when no command name is provided. Interactive mode can also use an external local game-master process:

./bin/golab --gm external \
  --gm-command /home/alice/projects/coolio-arena-master/bin/coolio-arena-master \
  --gm-interval 120 --gm-timeout 750ms

The external process receives one compact observation JSON on stdin and returns one event JSON on stdout. If the command fails or times out, golab falls back to the mock game master for that observation.

---

🎮 Controls

Action	Input
Pan view	Drag with left mouse button
Zoom in/out	Mouse wheel scroll
Pause simulation	Press P (config toggle)
Reset simulation	Press R
Cycle render mode	Press V
Save genome	Press G
Save map	Press M
Select god tool	Press 1-0
Use selected god tool	Left click or drag on board
Observe task path overlay	Hover over task-linked bots

Interactive saves are written as JSON under data/saves/genomes/ and data/saves/maps/. Render modes cycle through Normal, Genome, Health, Inventory, Colony, Task, Biome, and Pheromone.

---

🧠 Architecture Overview

core/
  bot.go         → Bot logic, genome, HP, energy
  colony.go      → Colony structure, task queues
  board.go       → Map and grid cell types
  genome.go      → Genome model and instruction logic 
  ...

game/
  game.go        → World loop, bot stepping, controller handling

ui/
  ui.go          → OpenGL rendering and input handling

• Bots have an instruction pointer, genome matrix, HP, and inventory
• Tasks are delegated by colonies: connect positions, maintain links, etc.
• Board is a grid of cells with typed content (walls, food, controller, etc.)
• Game loop runs bot logic and environmental updates on a tick-based basis

---