Voqal Tools are relatively simple to create but require creating custom yaml files. These yaml files could be generated
automatically by Voqal if there were a “Tool Mode”. This would allow developers to create Voqal Tools more quickly as
they could simply switch modes, have a conversation about a tool they want to create, loop through some conversational
trial and error to get the tool working, and then switch back to idle mode with a new tool available.
Help developers get started with Voqal by providing a guided tour of Voqal on installation. This would help developers
understand how to use Voqal and what it can do for them. It would also introduce developers to the configuration options
available to them.
Voqal could perform project-wide multi-step code refactoring with the help of a self-critique loop.
Currently, a developer can use Voqal to refactor code but that requires manually telling it everywhere to refactor.
With a self-critique loop, Voqal could automatically incorporate syntax and other IDE errors to autonomously
do large-scale refactoring. Then a developer could say something like "Add a toString method to all the DTO classes and make sure no personally identifiable information is included in the output" and Voqal handles it.
Adding support for knowledge base AI providers would allow someone to hook Voqal up to a private data source
and provide more accurate and relevant answers to user queries. For example, someone could hook up Voqal to Danswer
and ask Voqal if certain code has been changed, certain issues have been resolved, what issues are assigned to them, etc.
Adding support for speaker diarisation providers would allow a developer to ensure their Voqal Assistant only responds
to them and not other people in the room. This would be useful for developers working in shared spaces.
Assistant
Responses are generated using AI and may contain mistakes.