Assuming that one knows Feynman rules for QFT, what is the simplest way to generalize them for $T \neq 0$ case? What is the main difference? Can we just read them off from Lagrangian the same way as in QFT?
Since the partition function $Z$ in thermal quantum field theory is just $Z_{QFT}$ with substitution $t \rightarrow i\tau$, with $\tau$ being integrated from $0$ to $\beta$, as opposed to time $t$ being integrated from $-\infty$ to $\infty$, and with different boundary conditions for the fields (periodic/anti-periodic for bosons/fermions), can we use the same analogy for the generalization of the Feynman rules?