AgenticHub Release Notes_v0.5.2

1. Documents and Multimodal Input

Split View for Documents: After uploading a document, you can open a split view to read details while chatting with the agent alongside the source material.
Voice Input: Capture requirements by voice; they are shown as text in the conversation so you can start tasks quickly.

Design Skills: Dedicated skills for design workflows, with guidance better aligned to design-related tasks.
Pre-Sales Agent: Specialized agent capabilities for pre-sales and customer follow-up, improving logging and collaboration in structured processes.

Global Credentials: Manage access credentials centrally on the platform instead of re-entering them across features, reducing configuration overhead.
Hosting and Skills: Improved the end-to-end skill experience with GitLab; fixed cases where sandbox-related skills did not appear correctly in the UI.

Sandbox and First Launch: Fixed duplicate runtime environment creation in some scenarios; improved first-session environment preparation and reduced variable wait times.
Resources and Messaging: Better instance-creation resource availability in some regions; clearer, more noticeable prompts when environment capacity is tight so you can decide what to do next.
Sessions and Responsiveness: Refined how sessions bind to runtime environments to reduce inconsistent states; continued improvements to Skill loading and overall interaction smoothness.

Click Version List to view all historical versions.