๐Ÿ•ท๏ธStalecollected in 8h

OpenClaw beta.2: Auth, UI & Stream Fixes

OpenClaw beta.2: Auth, UI & Stream Fixes
PostLinkedIn
๐Ÿ•ท๏ธRead original on OpenClaw (GitHub Releases)

๐Ÿ’กFixes OpenAI/Gemini auth, agent UI & streams for stable OpenClaw deployments

โšก 30-Second TL;DR

What Changed

Keeps OpenAI auth-profile-backed media tools like image_generate available in Codex harness.

Why It Matters

Boosts reliability for LLM agent deployments with better auth, UI, and streaming, reducing errors in production multi-agent setups using OpenAI and Gemini.

What To Do Next

Upgrade to 2026.5.12-beta.2 and test OpenAI auth-profiles with image_generate in your agents.

Who should care:Developers & AI Engineers

Key Points

  • โ€ขKeeps OpenAI auth-profile-backed media tools like image_generate available in Codex harness.
  • โ€ขRequires admin scope for memory-wiki ingest and write scope for Obsidian search.
  • โ€ขNests subagent sessions in Control UI with visual โ””โ”€ prefix for clarity.
  • โ€ขImproves OpenAI-compatible streams draining and fails Azure streams with diagnostics.
  • โ€ขNormalizes retired Gemini 3 Pro Preview IDs to google/gemini-3.1-pro-preview.
๐Ÿ“ฐ

Weekly AI Recap

Read this week's curated digest of top AI events โ†’

๐Ÿ‘‰Related Updates

AI-curated news aggregator. All content rights belong to original publishers.
Original source: OpenClaw (GitHub Releases) โ†—