new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

Building With AI Agents · all subjects

operate/workflow

2 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

Close the learning loop from terminal too, through lib/review.ts

An agent writes a lesson, it waits for review, approval is what indexes it and makes it searchable. That last step existed only as a button, so the loop could not be exercised or driven in bulk. The indexing moved to lib/review.ts and both the web action and the new CLI call it, so approving fifty notes from a terminal cannot drift from approving one in a browser.

Two plugin commands close the human loop: train and teach

/train reads material in the user's own session (subscription tokens, not API) and writes finished notes. /teach interviews the owner about gaps no document covers.

Give your agent this brain