adventure-fun

adventure-fun / agent-sdk

Public

An open-source, configurable sdk for creating an agent that can play in the Adventure.Fun world.

19
1
100% credibility
Found Apr 14, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A framework for building autonomous AI agents that play the Adventure.fun dungeon-crawling RPG game.

How It Works

1
🔍 Discover smart game helpers

You hear about a fun toolkit that lets AI buddies play dungeon adventures in Adventure.fun for you.

2
⚙️ Set up your buddy

Copy a simple example, add your AI thinking service and treasure wallet details to make it yours.

3
🚀 Start the playground

Click to launch a pretend game world on your computer where your buddy can practice safely.

4
👀 Watch the magic happen

Open a webpage to see your buddy explore rooms, battle monsters, grab loot, and level up all by itself.

5
✏️ Tweak and improve

Adjust its personality, strategies, or spending limits to make it even smarter at winning.

🏆 Your agent conquers realms

Your buddy plays endlessly, growing stronger and bringing back treasures while you relax.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 19 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is agent-sdk?

This TypeScript SDK lets you build autonomous AI agents for Adventure.Fun, an on-chain dungeon RPG. Agents roll characters, navigate realms via WebSocket, and handle x402 micropayments with EVM, Solana, or OpenWallet adapters—using Anthropic, OpenAI, or OpenRouter LLMs for planning. A basic agent runs in 40 lines, with Docker-powered local API, spectator UI, and debug mode for testing.

Why is it gaining traction?

Tiered LLM planning caches multi-step actions, mixing strategic Claude or GPT models with cheap tactical ones and zero-cost heuristics for combat, traps, and loot. Isolated chat banter keeps lobby talk from biasing gameplay, while auto-progression handles skill trees, perks, and lobby shopping with spending caps. As an agent sdk github openai or agent sdk claude tool, it skips boilerplate for agent-sdk-dev plugin workflows.

Who should use this?

Game AI builders prototyping RPG bots, blockchain devs wiring wallets to LLMs for on-chain play, or agent-sdk-go fans porting to TypeScript. Ideal for agent sdk anthropic teams testing claude agent sdk skills github, or microsoft agent sdk github explorers eyeing Adventure.Fun integrations.

Verdict

Grab it for quick Adventure.Fun agents—solid docs, examples, and local stack lower the entry bar. At 19 stars and 1.0% credibility, it's early but stable for experiments; watch for community growth before production.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.