OpenHands, an AI coding agent platform, just shipped version 1.43.0 with a handful of new experimental features aimed at improving both onboarding and chat experience. If you use OpenHands to automate workflows or run coding agents in your stack, this release tweaks how you set up automations and interact with conversations, with several changes marked as experimental and surfaced through the new feature flags page.
The standout addition is the ability to create automations directly from the setup wizard (feat(automations): create automations enabled from the setup wizard). This means users can now configure automations earlier in their onboarding flow, rather than circling back after initial setup.
Experimental features in chat and settings
A new Features page now exposes experimental user feature flags. This lets users toggle early-stage features for themselves, and the first one to land is conversation reactions (feat(settings): Features page for experimental user feature flags, with Conversation reactions).
In the chat itself, two experimental features are now available:
- Decision cards can be rendered directly in conversation (
feat(chat): render experimental decision cards). These cards are meant to help users and agents make choices or surface key branches in a workflow. - Agent message edits are now visible in the chat history (
feat(chat): show the agent's message edits (experimental)). This helps track when the agent has updated a message, which is useful for debugging or understanding agent behaviour over time.
A small but useful UI change: automation run rows now show a conversation indicator icon, making it easier to track which runs are linked to which chats.




