One context, shared by every step.
Context Layer is a shared layer that keeps goals, sources, permissions, and next steps together, so people, apps, and agents all work from the same live context, carrying its proof and its boundaries.
01 / One layer, many consumers
The context stays put. The consumer changes.
Pick who is working. The shared layer below stays the same. Each consumer reads the same goals, sources, permissions, and next steps.
02 / What the layer holds
Four things, always together.
Goals
What you are trying to achieve, stated once and kept current as the work moves.
Sources
The evidence behind the work, linked in place so every step can trace back to it.
Permissions
What each actor may read, write, or trigger: boundaries that travel with the context.
Next steps
The move that comes next, so anyone (or any agent) can pick up without a re-brief.
FAQ
Questions, answered.
What is Context Layer?+
Context Layer is a shared layer that keeps goals, sources, permissions, and next steps together. People, apps, and agents all act on the same live context, and it carries its proof and its boundaries rather than a stale copy.
Why does shared context matter for agents?+
When reasoning happens in one place and execution in another with mismatched context, risk spikes at the handoff. A shared layer means an agent inherits the goal, the evidence, and the permissions together, so it acts with the same picture you have.
How is it different from a database or a doc?+
A doc is a snapshot; a database is storage. Context Layer keeps the four facets structured and linked as working context: current, permissioned, and portable across the apps and agents that consume it.
When can I use it?+
Context Layer is pre-launch. Join the waitlist to build on it early. It is part of Take Interest, so control and privacy are built in: you decide what is shared and with whom.
Keep the context together.
Context Layer is pre-launch. Join the waitlist to build agent workflows on shared, proof-carrying context.