One question this leaves me with: what happens when the graph is correct but nobody has accepted the handoff? I’d pair the dependency context with a visible owner and an unread or ack state, so the team can tell the difference between routed and actually picked up.
Good question, and I think it's pointing at a real seam. Riftmap's job stops at making the graph correct and legible: here's the blast radius, here's what's downstream, here's who/what it touches. Whether that's been accepted (routed vs. actually picked up, as you put it) is a coordination state that lives one layer up from the context itself. We deliberately don't own that state machine today; the context and the acceptance-of-context feel like two different primitives, and conflating them is how you get a graph that's technically right but operationally dead. Curious how you're thinking about the owner/ack layer on your end? That split (context underneath, coordination on top) seems like where a lot of the interesting design questions are.
Yeah, I think the claim has to be explicit. Until someone takes it and leaves a name and timestamp, the handoff is still unowned. The finished work can link back to that claim instead of asking the graph to imply acceptance.
Alook
Share your agents with people you trust. Bring AI agents running on your machine into a shared room, give your team a way to collaborate with them directly — a Discord-style workspace.
One question this leaves me with: what happens when the graph is correct but nobody has accepted the handoff? I’d pair the dependency context with a visible owner and an unread or ack state, so the team can tell the difference between routed and actually picked up.