file-reading

Routing uploaded PDFs to a dedicated parsing workflow prevents binary garbage by using pre-read inspection tools like pdfinfo, ensuring the model reads the correct format instead of raw binary data.

The best way to parse large CSV files without context flooding is size-aware reading, which uses lightweight inspection tools like stat and wc to sample contents before loading the full spreadsheet into the model's context.

File routing for uploaded documents works through extension-based dispatch, inspecting the file path under /mnt/user-data/uploads to select the appropriate parsing workflow based on the specific file type and format.

You can safely process uploaded archives by listing their contents first without auto-extraction. This archive safety precaution prevents unexpected file overflows unless explicit extraction is requested by the user.

Yes, this file routing approach handles Office formats by applying dedicated hand-offs to specialized skills for deeper docx, xlsx, and pptx processing, preventing the model from overloading a single workflow.