Skip to content

README.md

Latest commit

 

History

History
129 lines (98 loc) · 4.38 KB

README.md

File metadata and controls

129 lines (98 loc) · 4.38 KB
Raw

Memanto Memory Plugin for NeMo Agent Toolkit

Memanto is an open-source long-term memory layer for AI agents. The nemo-agent-toolkit-memanto plugin registers Memanto as a memory backend, _type: memanto, that works with the automatic memory wrapper agent and with the add_memory, get_memory, and delete_memory tools.

  • One Memanto agent per user. Each user_id maps to its own Memanto agent, so recall never crosses users.
  • Typed memories, not transcripts. Conversation turns saved by the automatic memory wrapper are distilled into facts, preferences, and decisions. Small talk stores nothing.
  • Explicit memories stored as written. Items with memory text, such as those from the add_memory tool, are stored as is, with their tags.

Installation

pip install nemo-agent-toolkit-memanto nvidia-nat-langchain

Get a Moorcheh API key at console.moorcheh.ai and set MOORCHEH_API_KEY in your environment:

export MOORCHEH_API_KEY=<YOUR_MOORCHEH_API_KEY>

Check that NeMo Agent Toolkit discovers the plugin:

nat info components -t memory -q memanto

Automatic Memory

Wrap any agent in the automatic memory wrapper agent. Each user message is saved, relevant memories are retrieved before the inner agent runs, and the agent's reply is saved. The user identity comes from the NeMo Agent Toolkit runtime.

memory:
  user_memory:
    _type: memanto

llms:
  nim_llm:
    _type: nim
    model_name: meta/llama-3.3-70b-instruct

functions:
  current_datetime:
    _type: current_datetime
  my_react_agent:
    _type: react_agent
    llm_name: nim_llm
    tool_names: [current_datetime]

workflow:
  _type: auto_memory_agent
  description: Assistant that remembers each user
  inner_agent_name: my_react_agent
  memory_name: user_memory
  llm_name: nim_llm
  search_params:
    top_k: 5

Save the configuration as config.yml, set NVIDIA_API_KEY, and run two turns:

nat run --config_file config.yml --input "I'm vegetarian and I live in Toronto."
nat run --config_file config.yml --input "Suggest a restaurant for me."

The first turn stores memories such as "User is vegetarian." and "User lives in Toronto." The second turn retrieves them before the agent answers.

Memory Tools

To let the agent decide what to remember, give it the memory tools:

memory:
  user_memory:
    _type: memanto

functions:
  add_memory:
    _type: add_memory
    memory: user_memory
    user_id: alex
  get_memory:
    _type: get_memory
    memory: user_memory
    user_id: alex
  delete_memory:
    _type: delete_memory
    memory: user_memory
    user_id: alex

workflow:
  _type: react_agent
  llm_name: nim_llm
  tool_names: [add_memory, get_memory, delete_memory]

A fixed user_id suits a single-user deployment. For multiple authenticated users, set user_id_resolver instead. Refer to Memory for details.

Configuration

Field Default Description
api_key MOORCHEH_API_KEY Moorcheh API key.
agent_prefix nat Prefix of the per-user agent IDs. Use a different prefix for each application so they never share memories.
max_memories 20 Maximum memories extracted from one conversation, from 1 to 100.

More Resources