Explanation

NEXUS is built around a handful of concepts that do not exist in a plain terminal workflow. Each page in this section takes one of them, starts from the command-line mental model you already have, and explains what NEXUS adds on top of it — what problem the addition solves, what work pattern or tool it replaces, and where its honest limits are.

These pages explain; they do not instruct. When you want to do something, the guides for each surface are linked from every page.

The concepts

• Agents — a terminal session that keeps its name and instructions between launches.
• Memory — an instructions file the agent writes to itself.
• Neural Map — a log of what your agents learned, drawn as a graph.
• Mesh — piping one agent's output into another's, made asynchronous and persistent.
• Teams — several agents pointed at one task, without the shell script.
• Task Board — a to-do list the whole team shares, where "done" has to be proven.
• Room — a live cockpit for everything above.

How to read this section

The order above is deliberate. Agents are the base unit; memory is what makes an agent worth keeping; the Neural Map is how you audit that memory; the mesh is how agents reach each other; teams and the task board organize that reach around a goal; the Room is where you watch all of it. Reading front to back gives you the whole model, but every page stands alone.

Each page ends with what the concept is not. NEXUS is local-first: nothing here syncs your terminals to a cloud, delegation between agents is depth-capped, and an agent can only touch what its capability grants allow. Those limits recur across the section because they are design decisions, not gaps.