Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.memanto.ai/llms.txt

Use this file to discover all available pages before exploring further.

​
memanto upload

Upload a file directly into the active agent’s memory namespace. The file content is processed and embedded by Moorcheh, making it instantly searchable via memanto recall. 
memanto upload FILE_PATH
Arguments:
  • FILE_PATH - Path to the file to upload (required)
Supported Formats: .pdf, .docx, .xlsx, .json, .txt, .csv, .md Maximum File Size: 5 GB Requirements:
  • An active agent session is required before uploading
Examples: Upload a PDF report: 
memanto upload ./quarterly-report.pdf
Upload a CSV dataset: 
memanto upload /data/customers.csv
Upload a markdown knowledge base: 
memanto upload ./knowledge-base.md
Output: 
Uploading quarterly-report.pdf (2.4 MB)...
✓ File uploaded successfully

  File:      quarterly-report.pdf
  Size:      2.4 MB
  Namespace: memanto_agent_customer-support
  Time:      3.2s

File content is now searchable via 'memanto recall'.
Recall Results from Uploaded Files: When recalling memories, file-sourced content is visually distinguished from manually stored memories: 
Found 3 memories:

1. Q3 revenue increased 18% year-over-year  · file upload · summary
   Confidence: 0.95
   Created: 2025-03-30 10:15:00 UTC

2. Customer churn rate dropped to 4.2%  · file upload · chunk
   Confidence: 0.92
   Created: 2025-03-30 10:15:00 UTC

3. Customer prefers email communication  · memory
   Confidence: 0.98
   Created: 2025-03-26 09:00:00 UTC
Notes:
  • Requires an active session (memanto agent activate)
  • Files are processed asynchronously by Moorcheh (embeddings generated server-side)
  • Uploaded content is stored in the cloud namespace, not as local memories
  • Unsupported file types are rejected with a clear error message