Task reflection SPI and in-memory implementation for AgentEnsemble — enables self-optimizing prompt loops
Version3.5.1compared with

Version Details

No dependency changes compared to the selected version.