Open-Source Agent-Based Learning Software
An open-source personal knowledge management tool that uses AI agents to automatically construct knowledge topology graphs, generate documents, and create exam questions, forming a closed learning loop. It supports Windows only.
Timeline (last 90 days)
The software currently only provides a Windows version, and the installation package can be downloaded from the Release page of the GitHub repository.
If an API Key is filled in, the provider configuration is automatically generated after creating a user, and no manual setup is needed afterward.
After a goal is created, the system generates an empty topology graph; users can start a task by dragging in documents or outlines, or by entering requirements in the Agent chat area.
Each time the Agent writes content, it obtains authoritative articles as supporting evidence through Web search, and provides reference materials and authoritative sources at the end of the document.
The note feature is mounted as a child node of a document, and users can create one by dragging a note onto a specified document.
A super document is an interactive HTML file that can call functions of the in-app environment, and is mainly used for experimental interaction.
The exam paper feature is designed as an addable node; the user drags an exam paper node under the document to be examined, the Agent asks for the number of exam questions, and custom question counts (such as 30, 40, or 100 questions) are supported.
Clicking "Start Exam" pops up a window, and the entire main application window turns black and is covered by a mask to prevent cheating.
After answers are submitted, the Agent enters the grading process; once grading is complete, an answer record appears under the exam paper node, including the graded status, exam analysis, and explanations, and there is also a screenshot detection feature.
Currently the software only supports Windows and the DeepSeek API, which limits cross-platform users and users of other models.
All Facts (10)
The software currently only provides a Windows version, and the installation package can be downloaded from the Release page of the GitHub repository.
50%UnverifiedIf an API Key is filled in, the provider configuration is automatically generated after creating a user, and no manual setup is needed afterward.
50%UnverifiedAfter a goal is created, the system generates an empty topology graph; users can start a task by dragging in documents or outlines, or by entering requirements in the Agent chat area.
50%UnverifiedEach time the Agent writes content, it obtains authoritative articles as supporting evidence through Web search, and provides reference materials and authoritative sources at the end of the document.
50%UnverifiedThe note feature is mounted as a child node of a document, and users can create one by dragging a note onto a specified document.
50%UnverifiedA super document is an interactive HTML file that can call functions of the in-app environment, and is mainly used for experimental interaction.
50%UnverifiedThe exam paper feature is designed as an addable node; the user drags an exam paper node under the document to be examined, the Agent asks for the number of exam questions, and custom question counts (such as 30, 40, or 100 questions) are supported.
50%UnverifiedClicking "Start Exam" pops up a window, and the entire main application window turns black and is covered by a mask to prevent cheating.
50%UnverifiedAfter answers are submitted, the Agent enters the grading process; once grading is complete, an answer record appears under the exam paper node, including the graded status, exam analysis, and explanations, and there is also a screenshot detection feature.
50%UnverifiedCurrently the software only supports Windows and the DeepSeek API, which limits cross-platform users and users of other models.
50%