[Release] openclaw/openclaw v2026.4.12: openclaw 2026.4.12

OpenClaw v2026.4.12 is a comprehensive quality release emphasizing plugin loading reliability, memory system improvements, and expanded local model support. Key additions include an Active Memory sub-agent plugin, local MLX speech provider for Talk Mode, LM Studio provider for self-hosted models, and enhanced Feishu setup workflows. The release also addresses critical security fixes, dreaming/narrative stability, and improves gateway lifecycle management and Matrix client streaming support.

Key Points

• •Active Memory plugin enables automatic context recall in ongoing chats without manual 'remember' commands, with configurable modes and transcript persistence for debugging
• •New local MLX speech provider for Talk Mode with explicit provider selection, interruption handling, and system-voice fallback for macOS
• •Bundled LM Studio provider adds runtime model discovery and memory-search embeddings for local/self-hosted OpenAI-compatible models
• •Plugin loading narrowed to manifest-declared needs with centralized manifest-owner policy to prevent unrelated plugin runtime activation
• •Local `openclaw exec-policy` command synchronizes tool execution config with approval files, including rollback safety and conflict detection
• •Dreaming system hardened with managed heartbeat event consumption, improved narrative cleanup, and prevention of self-ingestion of own transcripts
• •Memory/wiki system preserves Unicode characters in slugs and caps filename segments to prevent path overflow issues
• •Gateway `commands.list` RPC enables remote clients to discover runtime-native, text, skill, and plugin commands with serialized argument metadata
• •Security hardened with busybox removal from safe bins, empty approver list prevention, shell-wrapper detection, and placeholder credential validation
• •Matrix client support enhanced with MSC4357 live markers for draft preview sends/edits enabling typewriter animations

Workflow Diagram

Concepts

Artifacts (4)

openclaw exec-policy show
openclaw exec-policy preset <preset-name>
openclaw exec-policy set <config-path>

{
  "plugins": {
    "active-memory": {
      "enabled": true,
      "context_modes": ["message", "recent", "full"],
      "verbose_inspection": true,
      "transcript_persistence": true,
      "prompt_overrides": true,
      "thinking_overrides": true
    }
  }
}

{
  "models": {
    "providers": {
      "lm-studio": {
        "enabled": true,
        "runtime_model_discovery": true,
        "stream_preload": true,
        "memory_search_embeddings": true,
        "onboarding": true
      }
    }
  }
}

{
  "models": {
    "providers": {
      "openai-compatible": {
        "request": {
          "allowPrivateNetwork": true
        }
      }
    }
  }
}