# ChangeLog ## v0.7.1 - 2026-02-25 ### Enhancements: - feat(settings): add `settings_set_global_callbacks()` function to allow custom settings backend implementation - feat(settings): implement settings callbacks using NVS helper service (`get_key_value`, `save_key_value`, `erase_keys`) - feat(agent): simplify `on_interrupt_speaking()` - remove delayed task and directly call `reset_interrupted_speaking()` ## v0.7.0 - 2026-01-16 ### Initial Release - feat(repo): add XiaoZhi agent implementation for ESP-Brookesia ecosystem - feat(repo): implement OPUS audio codec support at 16kHz sample rate and 24kbps bitrate - feat(repo): support agent general functions: `InterruptSpeaking` - feat(repo): support agent general events: `SpeakingStatusChanged`, `ListeningStatusChanged`, `AgentSpeakingTextGot`, `UserSpeakingTextGot`, `EmoteGot` - feat(repo): support manual listening control via `on_manual_start_listening()` and `on_manual_stop_listening()` - feat(Kconfig): add `BROOKESIA_AGENT_XIAOZHI_ENABLE_AUTO_REGISTER` option for automatic plugin registration - feat(Kconfig): add `BROOKESIA_AGENT_XIAOZHI_GET_HTTP_INFO_INTERVAL_MS` for activation check interval - feat(Kconfig): add `BROOKESIA_AGENT_XIAOZHI_WAKE_WORD` for configurable wake word - feat(repo): integrate with agent manager framework for unified lifecycle management
41339bbf0e8093be4dca31344940cc6820e42aef
idf.py add-dependency "espressif/brookesia_agent_xiaozhi^0.7.1"