[Release] openclaw/openclaw v2026.6.8: openclaw 2026.6.8

OpenClaw v2026.6.8 is a major release focusing on improved channel delivery (Telegram/WhatsApp), enhanced agent and gateway recovery mechanisms, expanded provider/model support (GLM-5.2, Claude Haiku 4.5), better usage hooks with native footer rendering, and UI/mobile stability improvements. The release includes fixes for memory management, state handling, diagnostics, and numerous edge cases across channels, agents, and runtime systems.

Key Points

• •Telegram and WhatsApp delivery now support rich text formatting (tables, lists, blockquotes), preserved line breaks, and safer media boundaries
• •Agent and Gateway recovery improved across DM sends, media completions, auto-reply messages, archive fallback reads, and session identity handling
• •Added support for GLM-5.2 and Claude Haiku 4.5 models with normalized provider-qualified model IDs across OpenRouter and Google Vertex
• •Usage hooks now include native full footer rendering, default templates, fixed-decimal formatting, and credential-aware limits
• •UI/mobile improvements: workspace file collapsing, WebChat backscroll stability, interactive sidebar session picker, and iOS gateway reconnection
• •Memory and state management fixes: oversized embedding batch splitting, QMD memory search in transient mode, SQLite WAL avoidance on NFS volumes
• •Web search providers (Parallel Free, DuckDuckGo, Ollama) changed to explicit opt-ins instead of automatic selection
• •CLI-backed sessions now support `/btw` command with proper usage-error classification
• •Enhanced error handling for OpenAI reasoning signatures, Anthropic thinking signatures, and tool schema validation
• •Improved diagnostics with better partial-count handling, warnings for broken templates, and nested embedding fetch failure surfacing

Workflow Diagram

Concepts