hermes - 💡(How to fix) Fix [Feature]: LLM agent orchestrator [1 participants]

Problem or Use Case

not sure it s already planned but, for local setup, would be nice to have a feature allowing , on first prompt of a session, to let a default light llm auto load ( by analyzing prompt ) a dedicated agent with its own model.

As an example, i would like to use a light llm by default such as Qwen3 4b and if i prompt a coding request, it ll be replaced for the session by a more robust llm such as Qwen3.6 35 a3b because ia created a coding agent that use this llm.

on a /new it will go back to Qwen3 4b

The purpose of this is to be able to load specialized llm by task instead of loading huge model event for dumb tasks.

Proposed Solution

Having a default orchestrator agent with a light and dumb model which auto load a sub agent / replace itself, by analyzing first prompt.

i would be pleased to suggest a PR but i dont know where to start

Alternatives Considered

No response

Feature Type

New tool

Scope

None

Contribution

• I'd like to implement this myself and submit a PR

Debug Report (optional)