1. Add the ability to move files, create/delete directories, and rename directories in the sidebar via right-click. 2. Support Agent operations for moving files, creating directories, and modifying directories, but do not support deleting directories or files (deletion is not natively supported). 3. Support user prompt copying and rewriting, as well as AI reply message copying and retrying.

1. Added Agent Loop capability, no longer limited to traditional single-document block-level modifications. The Agent can call questioning tools, knowledge base retrieval tools, web search, and other tools based on user intent to modify and create multiple documents (deletion is not currently supported). 2. Added Diff popup, allowing users to automatically or manually apply Agent modifications to documents. 3. Added document upload functionality, supporting parsing of PDF, DOCX, MD, and TXT documents. Large text content is automatically saved as TXT files. Document content is always retained in the conversation context. 4. Added web search functionality, supporting Firecrawl, Tavily, Exa, and Zhipu search providers. Firecrawl offers a free monthly quota and API key applications are available at: https://www.firecrawl.dev/app 5. Uploaded images are uniformly saved to the image directory. When exporting documents, the image directory is included, and images are displayed using relative paths. 6. UI refactoring: Refactored the knowledge base, composition input box, and conversation area UI. Added toolchain functionality, allowing users to view called tools and execution steps during Agent execution. 7. Conversation context optimization: Tool calls, document and link parsing content, etc., are now included as Agent context.

1.Sidebar and main pages now remember scroll position and browsing state. 2.Knowledge base and creation pages now support deleting entire AI conversation history. 3.Optimized the creation Agent.

initial release