Agent DailyAgent Daily
releaseintermediate

[Release] openclaw/openclaw v2026.4.23-beta.4: openclaw 2026.4.23-beta.4

By steipetegithub
View original on github

OpenClaw v2026.4.23-beta.4 introduces significant enhancements to image generation across multiple providers (OpenAI, OpenRouter), adds optional forked context for subagents, implements per-call timeout support for generation tools, and includes configurable local embedding context sizes. The release also delivers numerous fixes for Codex harness routing, media handling across platforms (WhatsApp, Telegram, Slack), transcript replay behavior, and provider authentication flows to improve reliability and user experience.

Key Points

  • Image generation now works without API keys via Codex OAuth for OpenAI and through OpenRouter's image_generate tool, expanding provider accessibility
  • Agents can request provider-supported quality/format hints and pass OpenAI-specific parameters (background, moderation, compression) through the image_generate tool
  • Optional forked context for subagents allows child agents to inherit requester transcripts while maintaining isolated sessions by default
  • Per-call timeoutMs support for image, video, music, and TTS generation tools enables agents to extend provider timeouts for specific generation requests
  • Configurable memorySearch.local.contextSize (default 4096) allows tuning local embedding contexts for constrained hosts without patching
  • Codex harness now includes structured debug logging for embedded harness selection decisions, improving troubleshooting visibility
  • Fixed critical issues: native request_user_input routing, context-engine assembly failure redaction, WhatsApp first-run setup loading, and block streaming duplicate prevention
  • Media handling improvements across platforms: Telegram markdown image parsing, Slack group DM classification, WhatsApp media normalization, and WebChat image attachment preservation
  • Transcript replay behavior corrected to stop synthesizing missing tool results for OpenAI/Codex while preserving synthetic repair for other providers
  • Provider authentication hardened: OpenAI Codex OAuth routing, image generation auth validation, SSRF opt-in support for private networks, and model catalog recovery hints

Found this useful? Add it to a playbook for a step-by-step implementation guide.

Workflow Diagram

Start Process
Step A
Step B
Step C
Complete
Quality

Concepts