OpenClaw v2026.6.11 Stable: Reliability Over New Features

OpenClaw’s latest stable release, v2026.6.11, shipped June 30 with a clear message: the team heard the feedback, and this cycle is about polish over new capabilities.

The release targets dependability issues across nearly every supported channel. Highlights include:

Channel delivery reliability Misplaced replies, dropped attachments, and misrouted messages are fixed across Telegram, WhatsApp, Discord, Matrix, Google Chat, Feishu, iMessage, QQBot, and Mattermost. The release notes call out fixes for conversation tracking that was slipping across session changes and reconnections.

Heartbeat reasoning exposure When using reasoning-capable models for heartbeat checks, OpenClaw now surfaces the intended reply rather than exposing raw internal reasoning in Telegram, WhatsApp, and other channels. Thinking-mode messages remain available via opt-in.

Telegram UX improvements Progress-mode chats now clear old progress bubbles before new tool output or artifacts appear, keeping conversations readable. Telegram answers also now correctly stay attached to the user’s current question when quoting an earlier bot message.

QQBot group admin controls Group admins can now set how broadly slash commands are exposed — private-only commands redirect users to a private chat instead of being silently ignored.

Smarter Google Chat DM routing Newer Google Chat direct messages are no longer incorrectly treated as group conversations.

The v2026.6.11 comes after a rapid June release cycle that also brought native iOS and Android apps (June 29), a Skill Workshop for reusable agent workflows (June 3), and the NVIDIA SkillSpector collaboration for ClawHub supply-chain verification (June 1). The shift to reliability-focused work suggests the team is entering a consolidation phase after a period of aggressive feature expansion.