Every happy red team is the same, every unhappy red team is unhappy in its own way (Leo Tolstoy, probably)

A happy red team doesn't need much:

All of that generates data, most of that data is discarded. What isn't ends up in the report package the customer gets. Everything else? tribal knowledge, which walks away when a member leaves.

To avoid that, I like the micro-SOC approach: use blue team observability techniques (opentelemetry aggregators, logs and metrics collections around all your red team automation), so you can actually leverage it post-engagement.

Even if the initial leverage is just feeding a RAG pipeline to help with writing reports, that's time saved that can be used on the engagement itself.

And having a red team stack that could get ISO27001 certified as a standalone perimeter is pretty rad too! (you still have to write out your processes and policies though)