No dedicated adapter yet for Crew AI, but it could be still governed yet, as CrewAI runs its LLM call through LiteLLM, so point the model at Rikkernel by setting the base URL, this could happen without any code changes.
If you are running CrewAI for real, happy to open an issue for it and scope it.