frontend setting modal options to customize monolouge agent character traits and summary verbosity

[808vita]

What problem or use case are you trying to solve? Not able to set verbosity of summaries and other traits of the agent.

Describe the UX of the solution you'd like ability to customize the monologue agent's traits and summary verbosity from the frontend.

Do you have thoughts on the technical implementation?

• dropdown / input / sliders on the frontend inside settings modal.
• These can be passed along to the backend just like other selection settings ie. models , agent ,etc.
• on the backend those settings / custom prompt will be attached to the required prompts (action prompt , summary prompt)

Additional context

• introduces variations without breaking the agent.
• control the token usage to some extent
• can aid in: experimentations ,new agent building / discovery , prompt optimization , etc

[rbren]

I would suggest an architecture like this:

• Each agent provides a list of configurable parameters
• Each parameter needs a few fields:
  • name
  • description
  • type (number/string/text/enum)
  • default
• The UI ingests that list of parameters, and shows them in the settings modal when you select an agent
• Parameter settings are passed into the agent

This way it's very easy for agents to add new parameters, without also needing a UI change

[github-actions[bot]]

This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.

[github-actions[bot]]

This issue was closed because it has been stalled for over 30 days with no activity.