OpenClaw 2.0 has officially launched as the project’s largest update to date, incorporating contributions from 933 developers and merging more than 16,000 pull requests. The release represents roughly 50 percent of all pull requests ever merged into the open-source platform. The release arrived after a seven-week development cycle, marking a departure from the project’s prior cadence of 106 releases across 230 days.
The updated release revises core components of the system, including installation workflows, messaging architecture, memory handling, skills, automations, security layers, and native applications. In addition, the development team reworked both the internal codebase and deployment processes to accommodate an expanding contributor base, including 569 first-time contributors.
Simplified Installation and Setup
Initial setup now detects existing resources on a user’s machine to reduce configuration barriers. The setup workflow connects directly with existing API keys, local models, and subscriptions for ChatGPT or Claude. Users can initiate a basic setup and configure remaining preferences by interacting directly with the agent inside the chat interface.
The team rebuilt the browser application into a direct interface for starting conversations and monitoring active operations. Users can access ongoing workflows, view live agent actions, or manage settings without running external configuration scripts through the updated apps interface.
Task Automation Across Connected Services
The platform executes multi-step workflows across separate messaging and productivity tools. For example, a basic routine can monitor an email inbox for specific family notices and forward summary alerts directly to a Telegram account. Users can also instruct the agent to retrieve purchase records from email accounts and forward specific details through iMessage without manual searches.
These cross-service capabilities connect desktop software and local files with external artificial intelligence models. The automation framework handles routine retrieval and communication tasks while keeping user data under local control.
Architecture Changes in OpenClaw 2.0
Collaboration capabilities represent a major addition to the system through shared cloud sessions. Previously, team members could not join active workflows without resetting the agent’s context. The new shared workspaces allow multiple users to collaborate on live tasks, view project dashboards, and transfer ongoing sessions between team members while preserving operational memory.
The team used these collaborative workspaces internally to build the release itself. This structure allows engineers to hand off active development tasks directly within running environments on desktop computers.
Open-Source Development and Ownership
The project maintains an open-source model, allowing users to modify code and connect different model providers independently. The developers stated that the software is designed to let users control their own tools rather than relying on a single corporate provider or proprietary model structure.
