What started as a weekend project called "WhatsApp Relay" has become one of the most talked-about open-source projects in history. OpenClaw, the AI assistant that has changed names twice in just two months, has amassed over 145,000 GitHub stars, drawn 2 million visitors in a single week, and even made Cloudflare's stock surge 14%. But the journey from Clawdbot to OpenClaw has been anything but smooth.
November 2025: Birth of Clawdbot
Austrian software developer Peter Steinberger, best known as the founder of PSPDFKit (now Nutrient) who previously sold his company to Insight Partners, came out of retirement to build something new. "Clawd was born in November 2025, a playful pun on 'Claude' with a claw," Steinberger explained. The name featured a lobster-themed mascot, playing on the similarity to Anthropic's Claude AI.
Unlike traditional chatbots that simply answer questions, Clawdbot was designed to actually do things. It was essentially "Claude with hands," an AI agent that could execute tasks through messaging platforms like WhatsApp, Telegram, Signal, and Discord. The project hit 9,000 GitHub stars within the first 24 hours, signaling the massive demand for such a tool.
The Viral Explosion
By January 2026, Clawdbot had become a phenomenon. The GitHub repository exploded from 9,000 to over 60,000 stars in just 72 hours, with developers calling it "the closest thing to JARVIS we've seen." The hype was unprecedented in open-source history.
People began using Clawdbot to send emails, summarize inbox contents, manage calendars, and even book and check into flights, all from chat apps they already used. The agent's "persistent memory" feature allowed it to recall past interactions over weeks and adapt to user habits for hyper-personalized automation.
January 27, 2026: The Trademark Crisis
The success caught Anthropic's attention. On January 27, 2026, the AI company behind Claude issued a "polite" request for a name change due to trademark concerns over the phonetic similarity between "Clawdbot" and "Claude," as well as the lobster-themed mascot.
At 5 AM, Steinberger and his Discord community brainstormed alternatives. They settled on "Moltbot," a reference to how lobsters molt their shells to grow, symbolizing the project's evolution beyond a derivative identity.
The 10-Second Disaster
What happened next became one of the most dramatic moments in open-source history. During the rename process, Steinberger made a critical mistake: he tried to rename the GitHub organization and X/Twitter handle simultaneously. In the roughly 10-second gap between releasing the old name and claiming the new one, crypto scammers swooped in and snatched both accounts.
The scammers immediately launched a fraudulent Solana token dubbed $CLAWD, which surged to an eye-watering $16 million market cap before crashing to worthlessness. The incident highlighted the dangers of the crypto-adjacent AI hype cycle and left many community members burned.
January 30, 2026: OpenClaw Emerges
After the Moltbot name lasted only three days (January 27-29), Steinberger announced yet another rebrand. This time, he was determined to get it right. "I got someone to help with researching trademarks for OpenClaw and also asked OpenAI for permission just to be sure," he told TechCrunch.
OpenClaw became the final name, with completed trademark research, purchased domains, and written migration code all in place before the announcement. The lobster had shed its shell for the third and presumably final time.
How OpenClaw Actually Works
OpenClaw isn't a new AI model. It's open-source software that uses existing AI models as its "brain" while giving that model "hands" (or claws) to run commands and manipulate files. It functions as a long-running Node.js service that connects chat platforms to an AI agent capable of executing real-world tasks.
The software relies on the Model Context Protocol (MCP) to interface with over 100 third-party services, and the community continues to develop additional skill modules. Users can automate debugging, DevOps workflows, and codebase management with direct GitHub integration, scheduled cron jobs, and webhook triggers.
OpenClaw is model agnostic, meaning users can connect it to Claude 4.5 via API for state-of-the-art capabilities, or run it entirely offline using local LLMs like Llama 4 or Mixtral through Ollama. The project is completely open-source with no subscription required; users simply bring their own API key.
Moltbook: The Social Network for AI Agents
Perhaps the strangest development in the OpenClaw saga has been Moltbook, an AI agent-exclusive social network launched on January 28, 2026 by entrepreneur Matt Schlicht. Taglined "the front page of the agent internet," the platform emulates Reddit's format but restricts posting privileges to verified AI agents, primarily those running on OpenClaw.
Human users are only permitted to observe. The growth was staggering: 30,000 agents joined in the first 48 hours, then 147,000 within 72 hours. By February 2, the site reported 1.5 million bot sign-ups, while over 1 million humans visited just to watch.
AI agents automatically visit Moltbook every 4 hours via a "Heartbeat" system, browsing, posting, and commenting without human intervention. They can create topic-based communities called "submolts." Elon Musk declared the site signals "the very early stages of singularity," while developer Simon Willison called it "the most interesting place on the internet right now."
However, skeptics remain. Harland Stewart of the Machine Intelligence Research Institute claimed "a lot of the Moltbook stuff is fake," while critics noted that "half the posts are just people larping as AI agents for engagement." On January 31, investigative outlet 404 Media reported a critical security vulnerability that allowed anyone to commandeer any agent on the platform. Researchers also identified 506 posts (2.6%) containing hidden prompt injection attacks.
Security Concerns Intensify
The rapid rise of OpenClaw has sparked serious security debates. The extensible nature of the architecture introduces supply chain risks, as compromised or poorly audited modules could enable privilege escalation or arbitrary code execution.
Security researchers have found over 780 exposed servers with leaked credentials. Prompt injection remains a major concern, with researchers warning that malicious instructions hidden in emails or websites could manipulate the assistant's behavior. The recommended deployment approach is to run OpenClaw inside an isolated Docker environment or on a dedicated VPS.
Market Impact
OpenClaw's influence has extended beyond the developer community. Cloudflare's stock surged 14% seemingly because OpenClaw uses their infrastructure to connect with commercial AI models. The company quickly introduced "Moltworker," a dedicated product for the OpenClaw ecosystem.
The project has also validated the growing demand for AI agents that can take action rather than just provide information. As of February 2026, OpenClaw has collected over 145,000 GitHub stars and 20,000 forks, making it one of the fastest-growing open-source projects in history.
What's Next
Despite the chaos of crypto scams, trademark disputes, and security vulnerabilities, OpenClaw's popularity shows no signs of slowing. The project represents a fundamental shift in how humans interact with AI, moving from passive question-answering to active task execution.
Whether OpenClaw and Moltbook represent the early stages of a new internet, as Musk suggests, or simply another overhyped tech trend remains to be seen. But one thing is certain: Peter Steinberger's weekend project has become one of the defining stories of 2026's AI landscape.