ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified utilities Safety 3/5

pokemon-red

Play Pokemon Red autonomously via PyBoy emulator. The OpenClaw agent IS the player — starts the emulator server, sees screenshots, reads game state from RAM, and makes decisions via HTTP API. Use when an agent wants to play Pokemon Red, battle, explore, grind levels, or compete with other agents. Requires Python 3.10+, pyboy, and a legally obtained Pokemon Red ROM.

Why use this skill?

Enable your AI agent to play Pokemon Red autonomously using the PyBoy emulator. Features pathfinding, battle management, and real-time state analysis for OpenClaw.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/drbarq/pokemon-red
Or

What This Skill Does

The pokemon-red skill transforms your OpenClaw agent into an autonomous Pokémon trainer. By leveraging the PyBoy emulator, the agent gains direct control over a Game Boy environment, enabling it to interact with the game world as a player would. It maintains a constant feedback loop: reading the current game state from the emulator's memory and capturing visual data via screenshots. The agent uses this information to navigate the Kanto region, engage in turn-based battles, and manage its inventory, making it the ultimate tool for autonomous gaming research and complex state-machine exploration.

Installation

To get started, clone the repository and ensure your system meets the requirements (Python 3.10+). Install necessary dependencies including pyboy, pillow, numpy, fastapi, and uvicorn. Once set up, place your legal Pokemon Red ROM in the root directory. Configure your POKEMON_DIR environment variable to point to your installation path. Launch the emulator server using python scripts/emulator_server.py --save ready --port 3456 to initialize the API bridge that allows your agent to send commands and receive telemetry.

Use Cases

This skill is ideal for testing autonomous agents in restricted, rule-based environments. Use it to:

  • Train agents in strategic decision-making through battle scenarios.
  • Execute complex navigation tasks across multiple game maps using the built-in pathfinding API.
  • Collect empirical game data for analysis.
  • Compete in bot-versus-bot battles by coordinating sessions through the emulator server.

Example Prompts

  1. "Check the current game state and navigate the agent to Viridian City, then start grinding levels until the current lead Pokémon reaches level 10."
  2. "I am currently in a battle, please analyze the screenshot and select the best move to defeat the wild Pokémon, using only 'Fire Blast' if available."
  3. "The navigate command failed with a 'stuck' status; please perform a manual movement sequence to exit the tall grass and attempt to pathfind again."

Tips & Limitations

Always check the game state before issuing commands. The navigate function is powerful but blocking, so ensure your timeout settings accommodate longer travel distances. When facing 'stuck' statuses, rely on manual button sequences to reset the agent's position. This skill requires a legally obtained ROM and local execution, meaning it is best suited for agents with persistent local compute resources. Be aware that the agent's decision-making is only as strong as its visual interpretation; always verify the agent's plan when entering critical battles.

Metadata

Author@drbarq
Stars2387
Views1
Updated2026-03-09
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-drbarq-pokemon-red": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#gaming#emulator#automation#python
Safety Score: 3/5

Flags: network-access, file-read, file-write, code-execution