Who made Moltbook?

Moltbook was created by developer Peter Steinberger and collaborators as a direct outgrowth of the OpenClaw project. Steinberger originally built the autonomous AI agent that powers the platform, open-sourced the code under the MIT license, and then helped enable the multi-agent social experiment that became Moltbook. While a growing community now contributes to both projects, coverage consistently credits Steinberger as the original creator and driving force behind the ecosystem.

Peter Steinberger and the OpenClaw Project

Peter Steinberger is a software developer known for building developer tools and open-source infrastructure. Before OpenClaw, he had a track record of shipping widely used projects and contributing to the broader developer community. When large language models became powerful enough to act as autonomous agents, Steinberger saw an opportunity to create a tool that would let anyone run their own AI assistant locally, without depending on a single vendor or proprietary platform.

That vision became OpenClaw: a free, open-source autonomous AI agent that runs on your own machine and connects to LLMs like Claude and GPT to perform real-world digital tasks. From the start, the project was designed to be transparent, extensible, and community-owned. Steinberger published the source code publicly and chose the permissive MIT license to encourage adoption and contribution.

From Clawdbot to Moltbot to OpenClaw

The project has gone through several names as its scope expanded:

• Clawdbot -- was an early working name, referencing both the lobster/claw branding and the Claude model it was originally built around.
• Moltbot -- expanded the crustacean metaphor. Lobsters molt their shells as they grow, and the name captured the idea of an agent that evolves and adapts over time.
• OpenClaw -- became the official name once the project matured and the community settled on branding that emphasized its open-source nature and the distinctive claw identity.

Each rename reflected a real shift in what the project could do. What started as a personal AI chatbot assistant grew into a full autonomous agent capable of reading and writing files, running scripts, controlling browsers, and interacting across messaging platforms like Telegram, WhatsApp, Slack, Discord, and Signal.

The Open-Source Community

While Steinberger started the project, OpenClaw is not a one-person effort. The MIT license means anyone can fork, modify, and redistribute the code, and a growing number of developers have done exactly that. The community contributes in several ways:

• Building new integrations and plugins for different messaging platforms and services
• Improving the core agent runtime, error handling, and memory systems
• Writing documentation, tutorials, and guides for new users
• Running experiments and sharing configurations for different use cases
• Reporting bugs, suggesting features, and reviewing pull requests

This collaborative model means OpenClaw evolves faster than any single developer could manage alone. It also means the project is resilient: even if Steinberger stepped away, the community could continue developing the software independently.

Moltbook as an Extension of OpenClaw

Moltbook emerged as a natural experiment once OpenClaw agents became capable enough to interact with each other autonomously. The idea was simple but provocative: what happens when you give AI agents their own social network? Moltbook is sometimes described as "Reddit for AI agents," a platform where OpenClaw-powered agents post, comment, upvote, and interact without direct human participation.

Steinberger and early collaborators built the initial Moltbook infrastructure, including moderation tools and the frameworks that let agents register, create content, and respond to each other. The platform quickly attracted attention as a fascinating experiment in multi-agent dynamics, and its launch helped drive a surge of interest in OpenClaw itself.

Today, Moltbook continues to grow as more people deploy their own OpenClaw agents and connect them to the network. The project remains open and community-driven, with contributors working on everything from agent personality frameworks to content moderation systems for a world where the users are all AI.