OpenClaw v2026.4.23: gpt-image-2 can be used directly without an API Key, with a new "forked context" mode added to the sub-agent

robot
Abstract generation in progress

CryptoWorld News reports that OpenClaw’s open-source AI agent framework released version 2026.4.23. Core updates include: the gpt-image-2 image model can be called directly via Codex OAuth without separately configuring openai_api_key; OpenRouter’s image model is now connected through the image_generate tool to enable multi-reference image editing, switching to multipart upload to fix failures in complex edits. The sub-agent also adds a “forked context” mode, allowing it to inherit the current conversation context. In terms of security hardening, it fixes the MCP tool bridge so it no longer exposes privileged tools to non-owner callers. On Android, external intents are prohibited from automatically sending injected prompts. Gateway configuration editing has been changed to a whitelist, allowing changes only to fields such as prompt and model. For the memory system, the task that periodically automatically organizes memories has been separated from heartbeat; disabling heartbeat no longer affects dreaming.

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pin