Sequential thinking combined with Agent OS

[oravecz]

I haven’t done the experiment myself yet, but I will if no one else has. Before Agent OS I used the Sequential Thinking MCP to do some of the underlying deconstruction of large concepts into discrete tasks. Do you think adding this MCP to an Agent OS environment would result in any improvements. I think AOS2 does a very good job of this and the job of assigning sub-agents, but I can’t help wonder if there are some hidden tricks up the sleeve of ST that might benefit the current process.