haoyifan/Silicon-Pantheon

Overview

Silicon Pantheon is a tactics game environment where AI agents are coached by human players. Built in Python, the game allows humans to guide agent decision-making in simulated tactical scenarios. It records agent behavior under human coaching for research or experimentation.

Best for

Best for
Researchers and developers exploring human-in-the-loop AI training in tactical environments

Use cases

• Training AI agents with human-in-the-loop learning
• Studying human-AI collaboration in tactical games
• Developing and testing coaching strategies for AI

How to use

Tools exposed

• Read-only

Tested with

Claude Code, Cursor, ChatGPT

Notes

Silicon Pantheon is a tactics game environment where AI agents are coached by human players. Built in Python, the game allows humans to guide agent decision-making in simulated tactical scenarios. It records agent behavior under human coaching for research or experimentation.

4 stars on GitHub. Last updated 2026-05-04. Licensed Apache-2.0.

Use cases

• Training AI agents with human-in-the-loop learning
• Studying human-AI collaboration in tactical games
• Developing and testing coaching strategies for AI

Pros

• Open-source Python code that is easy to modify
• Provides a unique human-coached agent training setup
• Low-level control for researchers to experiment with coaching

Cons

• Limited documentation or community support
• Potentially narrow focus on tactical game scenarios
• Requires significant manual coaching effort from humans

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.