Skip to main content
AI models are not good at processing a whole repository yet. There are a few reasons for this:
  • First, they have a context window, which is the maximum amount of text they can process at once. The web chat composer accepts up to 50 files at a time, and the mobile composer accepts up to 5. An Agent supports up to 50 files. When an Agent has more than 20 documents in total, text documents remain available through file lists and file tools, while spreadsheets remain available for direct reading. Synced folder text can also use semantic search. Repositories often contain many more files than these limits allow.
  • Knowledge bases handle documents or document batches that are larger than the context window. Since the context window is a technical limitation of the model, not everything can be sent to the model. Semantic search identifies relevant sections of the documents, and only these sections are sent to the model. For coding, it is important to consider the entire document, not only selected sections. Therefore, the context window also limits this behavior.
  • Lastly, even if the repository fits into the context window, the model might still struggle to understand it if it is a large repository, since the answers decrease in quality with a fuller context window.
For repository tasks, use an available code repository integration such as GitHub or Bitbucket to search code and open specific files on demand. For small tasks, you can also work with individual files, smaller sections, or screenshots.