OpenClaw 2026.4.12 Brings Active Memory Agent, MLX Speech, and Feishu Overhaul

OpenClaw 2026.4.12 dropped today, marking a broad quality release focused on plugin loading, memory reliability, new local model options, and platform integrations.

Key highlights:

  • Active Memory plugin — A new optional plugin that gives OpenClaw a dedicated memory sub-agent running before the main reply. It automatically pulls in relevant preferences, context, and past details without the user needing to say “remember this.” Configurable message/recent/full context modes, live /verbose inspection, and opt-in transcript persistence for debugging.

  • macOS MLX speech provider — Experimental local MLX speech for Talk Mode, with explicit provider selection, local utterance playback, interruption handling, and system-voice fallback.

  • CLI exec-policy commands — New openclaw exec-policy command with show, preset, and set subcommands for synchronizing tools.exec.* config with the local exec approvals file, plus hardening for node-host rejection and sync conflict detection.

  • Gateway RPC commands.list — Remote gateway clients can now discover runtime-native, text, skill, and plugin commands with surface-aware naming and serialized argument metadata.

  • Private network model requests — New models.providers.*.request.allowPrivateNetwork option for trusted self-hosted OpenAI-compatible endpoints.

  • Feishu setup improvements — A smoother path for Feishu integration.

The release was announced on GitHub at 12:35 UTC today.

← Back to News