Skip to main content

Documentation Index

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

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

​
Overview

Argent memory has three lanes that must all be healthy:
  1. MemU/Memo: lived experience (what happened)
  2. RAG Library: uploaded knowledge (what was studied)
  3. SIS: pattern consolidation (what was learned over time)

​
Plain English

  • If MemU is down, Argent stops capturing new personal facts reliably.
  • If SIS is down, Argent stops improving lessons over time.
  • If RAG is down, document-backed answers degrade.

​
Required Baseline Settings

SettingValue
MemU LLM providerollama
MemU LLM modelqwen3:30b-a3b-instruct-2507-q4_K_M
MemU think levellow
MemU timeout30000
Embeddings modelnomic-embed-text (Ollama)
nomic-embed-text is an embedding model, not a MemU extraction/chat model.

​
Golden End-to-End Test

1

Send a unique personal fact

Send a unique personal fact in chat.
2

Recall the fact

Ask Argent to recall that fact after 1-2 turns.
3

Test RAG

Ask a direct question from a known uploaded library document.
4

Verify both lanes

Confirm both lanes respond correctly (MemU + RAG).