News in Short

• OpenClaw has released version 2.0, its biggest update since launch.
• The open-source AI agent platform now offers shared cloud sessions, smarter automations, conversation search, interactive dashboards and stronger credential security.
• The update involved 933 contributors and more than 16,000 pull requests.

OpenClaw’s latest update touches almost every part of the platform, including installation, messaging, memory, models, skills, browser tools, native apps, plugins and security.

OpenClaw 2.0 Makes It Easier to Get Started

One of the biggest changes comes at the setup stage. OpenClaw 2.0 can make use of resources users already have, including existing ChatGPT or Claude subscriptions, API keys and local AI models.

The platform has also moved some configuration steps out of the initial setup. This means new users can get to their first conversation faster and configure additional options later through conversations.

OpenClaw has also rebuilt its browser experience. Users can now return to ongoing work, continue setting up their AI agent and follow tasks in real time.

Cloud Sessions Let Users Continue Work Across Devices

OpenClaw 2.0 expands sessions beyond the Gateway. Users can continue work on paired devices or cloud workers while moving a session’s workspace along with it.

The platform can also reuse machines and project seeds for subsequent cloud sessions. This could make OpenClaw more useful for users who switch between devices or need AI agents to continue working remotely.

Durable session progress cards provide another useful addition. They let users track ongoing work, including subagent activity and accumulated edits, even after they reload the web or native app.

Search Past Conversations and Build Interactive Dashboards

OpenClaw 2.0 adds search for past conversations. Users can look for exact words or phrases in visible conversation text and reopen the messages around a matching result.

The update also introduces widgets and dashboards inside chats. Users can add widgets to conversations, pin them to session dashboards and control the actions or network origins they can access.

OpenClaw also lets users export rendered views as images, giving the new interface a more interactive feel.

Automations Can Now Handle Recurring Tasks

The update makes recurring AI tasks more convenient. Users can approve an exact operation once and allow an automation to repeat it later.

This could be particularly useful for repetitive workflows where users want an AI agent to perform the same approved action without asking for permission every time.

OpenClaw 2.0 also expands media support and adds new native playback controls.

OpenClaw Adds Stronger Credential Protection

Security gets an important upgrade as well. Agents can now request credentials through a masked prompt without exposing the actual value in the chat or the AI model’s context.

This gives users more control when an agent needs access to credentials while reducing the risk of sensitive information appearing in conversations.

OpenClaw 2.0 Is a Major Step for AI Agents

OpenClaw describes version 2.0 as the largest release in its history, following nearly seven weeks of development. The scale of the update is also reflected in its development effort, with 933 contributors and more than 16,000 pull requests.

With easier setup, cross-device cloud sessions, searchable conversations, recurring automations, interactive dashboards and improved security, OpenClaw 2.0 moves beyond being simply an AI agent platform. The update focuses on making AI agents easier to manage and more useful for longer-running workflows.