OpenGenieMaster
OpenGenieMaster is an open-source framework that lets coding agents control robots, read physical feedback, and learn from past jobs for robotics developers.
What it is good for
- Letting a coding agent control and supervise several robots
- Resuming interrupted robot tasks from the next unfinished step
- Testing robot task workflows in a simulator before using real hardware
- Saving lessons from robot runs without retraining the model
At a glance
- Who it is for
- Setup difficulty
- Technical ยท Guided command-line setup
Level 4 of 5. You copy and paste commands into a terminal by following the instructions. Comfortable territory for vibe coders.
- Cost
- Free and open source
- Type
- Framework or SDK
- Tasks
- Works with
- Claude Code, Codex, MCP, Anthropic SDK, Kimi, DeepSeek
How to get started with OpenGenieMaster
- Install Python 3.10 or newer.
- Run
git clone https://github.com/eust-w/opengeniemaster.gitthencd opengeniemasterandpip install -e .. - Run
ogm demo --scenario restocking --rounds 3 --trials 6to try the simulated store demo. - Connect Claude Code with
claude mcp add opengeniemaster -- ogm mcp --scenario restocking --state-dir .ogm. - Ask the agent to read the
geniemaster_operatorprompt, then give it a goal.
From the project's own documentation. If a step does not match what you see, the official pages have the latest instructions.
Before you try it
- It has not been tested on real robots; all results come from a simulator.
- Real robot use needs an emergency stop, human takeover, and motion limits in place.
- The PyPI package is not published yet, so install from source.
- The built-in Claude agent needs the Anthropic SDK and API credentials.
Momentum
- Rank this week
- #4 in builder tools
- Best rank
- #4 week 46, 2026
- Stars gained last week
- +348
- GitHub stars
- 455
GitHub numbers as of Nov 16, 2026. Updated weekly.
Ranking history
Official links
Questions
- Is OpenGenieMaster free?
- Yes. It is free and open source (MIT license).
- Do I need to know how to code to use OpenGenieMaster?
- You do not need to write code, but setup means copying and pasting commands into a terminal by following its instructions.
- What does OpenGenieMaster work with?
- According to its own documentation, it works with Claude Code, Codex, MCP, Anthropic SDK, Kimi, and DeepSeek.
- How do I get started with OpenGenieMaster?
- 1. Install Python 3.10 or newer. 2. Run git clone https://github.com/eust-w/opengeniemaster.git then cd opengeniemaster and pip install -e .. 3. Run ogm demo --scenario restocking --rounds 3 --trials 6 to try the simulated store demo. 4. Connect Claude Code with claude mcp add opengeniemaster -- ogm mcp --scenario restocking --state-dir .ogm. 5. Ask the agent to read the geniemaster_operator prompt, then give it a goal.
- What should I know before trying OpenGenieMaster?
- It has not been tested on real robots; all results come from a simulator. Real robot use needs an emergency stop, human takeover, and motion limits in place. The PyPI package is not published yet, so install from source. The built-in Claude agent needs the Anthropic SDK and API credentials.
- What is OpenGenieMaster good for?
- Letting a coding agent control and supervise several robots, resuming interrupted robot tasks from the next unfinished step, testing robot task workflows in a simulator before using real hardware, and saving lessons from robot runs without retraining the model.
Similar agents to OpenGenieMaster
- OpenHands
Runs coding agents for software work rather than controlling robots.
- AgentRun
Focuses on cutting token costs and recovering interrupted agent runs, not physical robot control.
- Cambium
Makes agent work on a knowledge repository auditable and resumable, with no robots involved.
- Agency Orchestrator
Turns one sentence into a team of AI expert roles that produce plans, not robot actions.
- browser-use
Lets an agent operate a web browser instead of physical robots.
- Kestra
Kestra is an open-source orchestration platform that schedules and automates data, infrastructure, and AI agent workflows for technical teams.