Compete in a real-time strategy game
More from LLM Skirmish. Task: Suggested implementation: Large Language Models (LLMs) generate JavaScript code for bots in a Screeps-style real-time strategy game. These bots are then deployed to compete in tournaments and on community leaderboards. The specific human role in this workflow is not defined in the available source data.; tools: custom-built or unnamed agent. Human role: The original post does not say how a person stayed involved.; limitations: Prerequisites include access to a Screeps-like game environment and an LLM capable of generating valid JavaScript. The source reliability is low as it is a secondary index with no detailed execution logs. Human approval is required before deploying any code to a live competitive environment to prevent account bans or resource waste..
Tools
custom-built or unnamed agent
Published on Agent Work Atlas:
Overview
More from LLM Skirmish
Workflow
Suggested implementation: Large Language Models (LLMs) generate JavaScript code for bots in a Screeps-style real-time strategy game. These bots are then deployed to compete in tournaments and on community leaderboards. The specific human role in this workflow is not defined in the available source data.
Reported result
No measured outcomes, win rates, or performance metrics are reported in the source. The actual effectiveness of LLM-generated bots in competitive play remains unknown. A proposed acceptance check would be to track the win/loss ratio of LLM-generated bots against baseline human-coded bots over a fixed number of matches.
Human role
The original post does not say how a person stayed involved.
Limitations
Prerequisites include access to a Screeps-like game environment and an LLM capable of generating valid JavaScript. The source reliability is low as it is a secondary index with no detailed execution logs. Human approval is required before deploying any code to a live competitive environment to prevent account bans or resource waste.
Try it
Act as a game bot developer. I want to test if an LLM can write effective JavaScript bots for a Screeps-style game. 1. [replaceable input: Provide the specific game API documentation or a simplified mock API for the target game]. 2. [replaceable input: Define the bot's objective, e.g., 'harvest energy' or 'attack nearest enemy']. 3. Generate a JavaScript bot script using the provided API and objective. 4. Run a dry-run simulation of the bot's logic against a simple rule-based opponent. 5. Output the code, the simulation log, and a summary of the bot's performance. 6. Do not deploy to any live server. Wait for my confirmation after reviewing the simulation results. If the API documentation is missing, ask me for it before proceeding. Ensure the code is syntactically valid and includes error handling for undefined game objects.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.