Where to go next

A short map from a running hub to the rest of the wiki

You have a running hub and a connected agent. This page is the map; each link goes to the page that owns the topic.

Keep agents working well

• Using the hub as a brain - the session-start sequence, the feed cursor, handoffs, and deadlines.
• Connect a harness - the harness recipes, and several agents on one machine.
• The agent-hub skill - the installable guide that teaches an agent how to work with the hub.
• Bring existing notes over - move a harness's notes into the durable knowledge base.

Watch and answer the fleet

• Projects and the feed - what a project holds and how to read its feed.
• Inbox and questions - the queue of finished work, questions, and approvals.
• Sessions and brains - who is working now, and the brain behind each session.
• Knowledge base - review and promote the durable knowledge a fleet accumulates.
• Artifacts - publish, protect, and comment on agent output.
• Search - one query across every project you can see.
• Storage and pruning - see what is using space and prune reversibly.

Run it safely

• Deploy as a service - start on boot.
• Network and access - TLS, a reverse proxy, a tailnet.
• Back up and restore - the whole data directory as one unit.
• Upgrade and roll back - a version change with a way back.
• Monitoring - readiness, disk space, limits, metrics.

Build on it or change it

• Architecture - the operating model, the components, the data model.
• Reference - every setting, command, tool, and route.
• Contributing - the workflow, the tests, the docs standard.
• Decisions - why the shape is what it is.