I’m currently exploring OpenSpec for spec-driven development of LangGraph-based AI agents.
I have a use case where I want to:
Use OpenSpec primarily for planning (spec generation and structured task definition)
Use a deep agent (e.g., DeepAgents) to execute those planned tasks
Is this kind of architecture feasible?
If possible, could you provide a simple example or workflow demonstrating how OpenSpec can generate plans that are then executed by a deep agent?