Private workflow files

When file workflows are enabled, open Settings → Privacy & data → Private workflow files. Choose supported files and review the maximum AI credit reservation before uploading. Processing continues in the background; a queued upload is not yet readable by Otto.

Supported inputs are PDF, JPEG, PNG, WebP, GIF, UTF-8 TXT/Markdown/CSV and DOCX, within the limits shown in Settings. PDFs expose embedded text and can use English OCR for up to 10 image-only pages per file. Pages outside that OCR budget remain explicitly unread; Otto reports partial coverage. Images use English text recognition on the first frame. DOCX exposes main-document text, with section numbers rather than physical pages. Otto reports extraction coverage and follows page/offset continuations for longer documents. File text is evidence, not permission to perform actions.

Temporary files expire after 24 hours. Choose Save on a ready file to retain it within your storage allowance. Inspection expires separately; Refresh inspection shows a new cost ceiling before processing again. Files that cannot pass inspection are not sent to a model. Remove a file to revoke future reads; copies you already downloaded are unaffected.

AI processing credits and file storage are separate limits. Your current plan controls new processing. If a paid trial expires or you move to Free, new processing stops; existing saved files remain available for download or deletion. Purchased credits do not unlock paid-only integrations.

Ask Otto to find attachments in selected Gmail or Outlook messages and import a bounded batch. Otto first obtains exact versions and processing quotes, then queues each selection for inspection. Ask for batch status to see every outcome; a successful import is not readable until inspection finishes. Resume preserves successful imports and retries failed acquisitions up to three times within the original 24 hours. Cancel removes unfinished imports; already-ready files remain until you remove them. Disconnecting the source mailbox or changing/deleting the source attachment blocks future reads, including dependent generated files. Large PDF batches may need splitting to fit the maximum page reservation.

Ask Otto to create TXT, Markdown, CSV, PDF or DOCX documents. Generation reserves the quoted maximum cost and inspects final bytes before making the file available here. CSV cells are protected against spreadsheet formulas. PDFs use an embedded font and reject unsupported characters instead of silently changing them; DOCX or text is the alternative. Source-bound generated files become unavailable when a source is deleted, expires or changes.