> For the complete documentation index, see [llms.txt](https://hyperdust-foundation.gitbook.io/moss-ai-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://hyperdust-foundation.gitbook.io/moss-ai-docs/archived/game/ai-agent-knowledge-management/launch-rag-knowledge-base/how-to-create-rag-knowledge-bases.md).
# How to Create RAG Knowledge Bases
**1.Define the Scope of Your Knowledge Base**\
First, determine what information your knowledge base will include. This data can consist of:
* Business-related documents (e.g., whitepapers, research reports, industry analyses)
* Product manuals, user guides, FAQs
* Articles, blogs, forum posts, community discussions
* Code repositories, API documentation
* Any proprietary knowledge you want the agent to have.\
**2.Prepare Knowledge Base Data**\
To ensure the RAG model can effectively use the data, you need to organize the documents:
* Formatting: Currently, only plain text is supported. Remove images, special characters, and unnecessary line breaks. Hierarchical titles using numbers can be retained.
* Text Splitting: Long texts should be broken down into manageable knowledge snippets for efficient indexing.
* File Formats: Supports .doc and .txt files.
* File Size: You can upload multiple files, with a maximum file size of 2MB per file.
* Knowledge Base Updates: The knowledge base can be updated at any time, allowing for flexible additions or deletions.
\
