godot-monte-carlo-balancer — Claude Code skill: balance setup
The godot-monte-carlo-balancer skill audits Godot balance with Monte Carlo simulations, economy checks, and parameter search. Use it to find unfair levels, farm exploits, and curve problems.
LGPL-3.0 268 installs ★ 983 updated 2026-10-16
npx skills add thedivergentai/Game-Dev-Agentic-Skills --skill godot-monte-carlo-balancer -a claude-codeWhat godot-monte-carlo-balancer does
godot-monte-carlo-balancer builds a source-driven balance lab for Godot projects. It reads live values from Resources and source code, records the model in BALANCE_PLAN.md, simulates playstyles from AFK to pro, and reports win-rate and economy verdicts with confidence intervals.
The package includes a Rust template using rayon, PowerShell and shell launchers, the balance-lab CLI, CI snapshot comparison, and a stable --json field contract. Its commands cover inspect, simulate, career, mode, bruteforce, gen-level, and calibrate. Parameter search can emit .tres files when the project follows a Resource-first setup.
For a mobile game, the simulation matrix must include touch input models. A mouse-only check can mark a level acceptable while touch play remains too hard.
What it helps with
- audit win rates across playstyles
- spot unreachable shops and interest-curve cliffs
- detect farming exploits in the economy
- compare balance snapshots in CI
- search level parameters by brute force
Example prompts for the agent
- Audit this Godot game's balance, run
balance-lab inspect, and createBALANCE_PLAN.mdbefore simulating anything. - Run
balance-lab simulate --level 3 --style average --runs 1000, then compare the result with the pro playstyle. - Check career economy behavior with
balance-lab career --style casual --runs 200.
How to install godot-monte-carlo-balancer
1Skills CLI (Vercel)
npx skills add thedivergentai/Game-Dev-Agentic-Skills --skill godot-monte-carlo-balancer -a claude-code2Manually: put the files in place
~/.claude/skills/godot-monte-carlo-balancer/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/godot-monte-carlo-balancer/
1Skills CLI (Vercel)
npx skills add thedivergentai/Game-Dev-Agentic-Skills --skill godot-monte-carlo-balancer -a codex2Manually: put the files in place
.agents/skills/godot-monte-carlo-balancer/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add thedivergentai/Game-Dev-Agentic-Skills --skill godot-monte-carlo-balancer -a cursor2Manually: put the files in place
.cursor/skills/godot-monte-carlo-balancer/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install thedivergentai/Game-Dev-Agentic-Skills skills/godot-monte-carlo-balancer2Manually: put the files in place
.github/skills/godot-monte-carlo-balancer/SKILL.md1Skills CLI (Vercel)
npx skills add thedivergentai/Game-Dev-Agentic-Skills --skill godot-monte-carlo-balancer -a gemini-cli2Platform CLI
gemini skills install https://github.com/thedivergentai/Game-Dev-Agentic-Skills --path skills/godot-monte-carlo-balancerManually: put the files in place
.devin/skills/godot-monte-carlo-balancer/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add thedivergentai/Game-Dev-Agentic-Skills --skill godot-monte-carlo-balancer -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about godot-monte-carlo-balancer
How do I install godot-monte-carlo-balancer in Claude Code?
Run npx skills add thedivergentai/Game-Dev-Agentic-Skills --skill godot-monte-carlo-balancer -a claude-code, then check ~/.claude/skills/godot-monte-carlo-balancer/SKILL.md.
What does godot-monte-carlo-balancer do?
It extracts Godot game data, runs simulations for several playstyles, evaluates win rate and economy behavior, and searches for better level parameters.
Does a mobile game need a separate balance check?
Yes. The matrix must include touch input models for mobile games, since mouse-only results can leave touch play too difficult.