PulseAudio 2.0.25: a rule chain with a Play Audio node now saves and runs; before, it failed to compile on every node.
Google TTS 2.0.20: the Speak Text node is removed. It only prepared the audio and never played it, and it stayed silent when the same sentence came twice in a row. Use Generate Speech followed by Play Audio, with the audio file from input set to data.result.audio; repeated sentences play every time.
⚠️ Breaking Changes
Rule chains that use the Google TTS Speak Text node stop compiling after the update. Replace the node with Generate Speech followed by Play Audio.