[Release] openclaw/openclaw v2026.4.25-beta.2: openclaw 2026.4.25-beta.2

OpenClaw v2026.4.25-beta.2 delivers major upgrades across text-to-speech (TTS) with new providers and chat-scoped controls, plugin architecture optimization through cold persisted registry, comprehensive OpenTelemetry observability expansion, safer browser automation with CDP improvements, and hardened installation/update processes across all platforms. The release includes PWA/Web Push support in the control UI, per-agent and per-account TTS overrides, and deterministic plugin discovery that eliminates broad manifest scans.

Key Points

• •TTS system overhaul: `/tts latest` read-aloud, chat-scoped auto-TTS toggles (`/tts chat on|off|default`), per-agent and per-account overrides, plus 8 new providers (Azure Speech, Xiaomi, Local CLI, Inworld, Volcengine, ElevenLabs v3, and channel-specific support)
• •Plugin registry migration: Move from manifest-based discovery to cold persisted registry, eliminating broad scans and making plugin updates, repairs, and provider discovery deterministic and faster
• •OpenTelemetry expansion: Bounded low-cardinality telemetry across model calls, token usage, tool loops, harness runs, exec processes, delivery, context assembly, and memory pressure with signal-specific OTLP endpoint overrides
• •Browser automation hardening: Safe tab URLs in responses, iframe-aware role snapshots, CDP readiness tuning, headless one-shot launch mode, and `openclaw browser doctor --deep` for slow-host diagnostics
• •Control UI enhancements: PWA install support, Web Push notifications for Gateway chat, Crestodian first-run repair, full-TUI interactive setup, context mode selector, and shorter startup greeting
• •Cross-platform install/update security: Hardened processes for Windows, macOS, Linux, Docker with bundled plugin runtime deps, Node service restarts, LaunchAgent token rotation, and mixed-version gateway verification
• •Provider discovery optimization: Resolve provider ownership and catalog hooks from cold registry instead of manifest rescans; use Provider Index as final fallback for installable providers
• •Plugin hooks expansion: Add before-agent-finalize hooks, cron `jobId` context, bounded native permission fingerprints, and Codex MCP hook relay support
• •Diagnostics improvements: Prometheus bundled plugin with protected gateway scrape route, W3C traceparent propagation from trusted trace context, and bounded exporter health diagnostics without raw error text
• •CLI enhancements: `openclaw plugins registry` for explicit inspection, `openclaw browser start --headless` one-shot launch, `--background` flag for image generation, and faster `models list --all --provider` via plugin index

Workflow Diagram

Concepts

Artifacts (5)

# Per-agent TTS override example
agents:
  list:
    - id: agent-1
      tts:
        provider: azure-speech
        voice: en-US-AriaNeural

# Channel and account TTS override
channels:
  discord:
    accounts:
      account-123:
        tts:
          provider: elevenlabs-v3
          voice: custom-voice-id

# Chat-scoped auto-TTS control
messages:
  tts:
    enabled: true
    provider: azure-speech

# One-shot headless browser launch
openclaw browser start --headless

# Deep browser diagnostics for slow hosts
openclaw browser doctor --deep

# Image generation with background option
openclaw infer image generate --background "solid-color" --output-format png
openclaw infer image edit --background "gradient" --output-format jpeg

# Plugin registry inspection
openclaw plugins registry
openclaw plugins list
openclaw plugins doctor --fix

# Models list optimization
openclaw models list --all --provider openai

# Signal-specific OTLP endpoint overrides
OTLP_TRACES_ENDPOINT=http://localhost:4317/v1/traces
OTLP_METRICS_ENDPOINT=http://localhost:4317/v1/metrics
OTLP_LOGS_ENDPOINT=http://localhost:4317/v1/logs

# OpenTelemetry semantic convention stability opt-in
OTEL_SEMCONV_STABILITY_OPT_IN=gen_ai_latest_experimental

# Diagnostics configuration
diagnostics:
  otel:
    enabled: true
    exporters:
      - type: otlp
        protocol: grpc
        endpoint: http://localhost:4317
  prometheus:
    enabled: true
    plugin: diagnostics-prometheus
    gateway_scrape_route: /metrics

# Before-agent-finalize hook example
plugins:
  hooks:
    - type: before-agent-finalize
      plugin_id: custom-plugin
      handler: onFinalizeAgent
    
    # Cron job with jobId context
    - type: cron
      schedule: "0 0 * * *"
      plugin_id: scheduled-task
      jobId: daily-cleanup
      handler: onCronTick
    
    # Codex MCP hook relay
    - type: mcp-relay
      plugin_id: codex-integration
      mcp_server: codex-mcp
      handler: relayMcpCall

# Discord voice channel LLM override
channels:
  discord:
    voice:
      model: gpt-4-turbo  # Override default LLM for voice responses
      stt_provider: openai-whisper  # Keep existing STT
      tts_provider: elevenlabs-v3   # Keep existing TTS
      
# TTS/WhatsApp chat-scoped auto-TTS
channels:
  whatsapp:
    tts:
      enabled: true
      provider: azure-speech
      # Chat-scoped override via /tts chat on|off|default