Msty Claw 0.12.0: Keep Context Close
Release 0.12.0 brings stronger chat attachments, per-chat drafts, Search Lens, message trail navigation, clarifications, Msty Nexus integration, mobile polish, easier access to services, a cleaner composer, and better tray visibility on Linux.
Improved Conversation Context
In 0.12.0, agents are now better at preserving the working state around a conversation.
Attachments stay tied to the message they support. Draft text and pending attachments stay scoped to the chat where they started. Search results carry more source context. Long conversations get a lighter way to recover earlier messages.
Together, these changes reduce the hidden cost of switching between chats, files, sources, and long threads. Claw keeps more of the surrounding context intact, so you spend less time rebuilding the setup around the work.
This results in fewer interruptions between agent's thought, context, and the next useful step.
Files and Drafts Stay Together
Attachments are more reliable across normal chats, agent chats, and Msty Claw Go.
Documents, images, notes, screenshots, and other supported files stay connected to the message they support. Useful conversations often depend on material outside the chat, so that context needs to remain attached to the right turn.
Drafts now follow the same per-chat boundary. Unsent text and pending attachments are saved against the conversation where they started, so switching chats restores the correct draft instead of carrying it into another thread.
Search Lens
Search Lens brings a more visual experience to web search results by displaying relevant images alongside each source's title and description. Instead of scanning through blocks of text, you can quickly identify the most relevant results at a glance.
After all, we process visual information naturally. By combining images with traditional search results, Search Lens helps you recognize, compare, and find what you're looking for faster.
Clarifications and a More Reliable Mobile Experience
When Claw needs a little more information to complete a task, it now handles the process more smoothly.
In 0.12.0, we've refined the clarification flow with clearer validation, a simpler review experience, and better state management, so your progress is preserved while you answer follow-up questions.
On mobile, we've also focused on reliability where it matters most. Improvements to clarification screens, attachments, Android photo selection, and response cancellation make Msty Claw Go more dependable during everyday use.
Together, these updates make it easier to pick up where you left off, whether you're working from your desktop or your phone.
Everyday Improvements
0.12.0 also includes a number of quality-of-life improvements throughout Claw.
The empty chat screen now points you to services that haven't been configured yet, making it easier to discover and enable available capabilities.
Other improvements include:
- Easier access to Web Search, Related Search, Data Protection, Airlock, and Computer Use.
- The Install Services prompt now only appears when there are services available to install.
- A cleaner composer layout that better organizes attachments, extensions, and chat controls.
- A model picker that's always within reach without taking up unnecessary space.
- Improved Linux tray icon visibility across different themes and desktop environments.
These aren't major features, but they remove a lot of the little annoyances that come up when using Claw every day.
Built to Keep Your Place
This release is focused on continuity.
Drafts are saved per chat. Attachments stay with the conversations they're added to. Search Lens makes web results easier to scan. Clarifications are smoother, mobile is more reliable, and Msty Nexus is now integrated directly into Claw.
Update to 0.12.0 and take it for a spin.
Msty Claw Beta
Start using Msty Claw
Install Msty Claw in minutes, connect your workflows, and run autonomous tasks with privacy-first controls.
Free for personal use in beta.
Built For Real Work
- Privacy-first controls with local and online model options.
- Assign outcomes and let agents execute multi-step tasks.
- Use Memory Packs, Playbooks, and Tasks to stay consistent.