List view

• milestone 4: virtualised compute

  Provide a safe environment where LLMs can perform tasks

  No due date
  0% complete0 open0 closed

• milestone 3: agents enabled

  milestone 3: to be specified but has the goal of enabling basic agents to act and have executive function (i.e. goal setting)

  No due date

  •1/1 issues closed
  100% complete0 open1 closed

• milestone 2: parsing, tools, long-term memory, goal-setting

  milestone 2 is all about taking preparing to enable proper agents for this: - parsing: generalise the parsing responses from the models. focus on yaml as the format - tools: improve the implementation of tools so that it is more generic and better handles errors - long-term-memory: adding support for vector stores we can give models long-term memory - goal-setting: add executive functionality to help agents break down tasks

  No due date

  •2/2 issues closed
  100% complete0 open2 closed