Connect an assistant
By the end, your AI assistant is connected to your Glassy under its own name, and you know how to take its access away.
Connect an agent, answer what it asks, and write together.
By the end, your AI assistant is connected to your Glassy under its own name, and you know how to take its access away.
By the end, you can see every agent that works in your Glassy, what each one did this week, and what it keeps for itself.
By the end, you can answer an agent's question with one click, and know what happens next.
By the end, you can see an agent's work land as records you can check.
By the end, you can work on something with an agent: it drafts, you edit it in place and answer with a click, and it picks up from there.
By the end, you can drive Glassy from your agent's chat, with buttons that do one thing each.
By the end, you can tell at a glance what you wrote and what an agent wrote.
By the end, you know where your agents keep their own notes, how to look at them, and how to move one into your workspace or back.
Longer reads, without a video yet.
The five questions worth asking, how to scope them by source and date, and how to read a cited answer without trusting it blindly.
Backlinks, neighborhoods, shortest paths, orphaned notes and the hubs your thinking orbits — the questions text search cannot answer.
Durable memory that is a real note, recall scoped to the right agent, and session digests you can read — memory without a second database.
Give each assistant its own named key on your own machine, see what it wrote, and answer what it asks. The guide lives in Glassy Self-Host.