What it does
Run an offline agent loop whose local decision model checks whether to act, continue or stop without another language-model call.
Agent tooling
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
Run an offline agent loop whose local decision model checks whether to act, continue or stop without another language-model call.
Start by listing every choice your helper makes while doing a job. These choices might include picking a tool, choosing which file to view, or checking whether the task is complete. Write each step down as a simple multiple-choice question or a yes-or-no check.
A developer can then connect a smaller offline program to make those quick picks on your machine. This setup saves time and keeps your main writing AI focused entirely on drafting the final responses.