[Release] openclaw/openclaw v2026.6.8-beta.2: openclaw 2026.6.8-beta.2

OpenClaw v2026.6.8-beta.2 delivers significant improvements across messaging channels, agent recovery, provider/model handling, and UI stability. Telegram and WhatsApp now support richer message formatting with better delivery reliability, while agent and gateway recovery mechanisms are enhanced across multiple failure scenarios. The release includes expanded provider support (GLM-5.2, Claude Haiku 4.5), improved usage hooks with native footer rendering, and numerous fixes for memory, state diagnostics, and mobile UI flows.

Key Points

• •Telegram and WhatsApp channel delivery enhanced with structured rich text (tables, lists, blockquotes), preserved line breaks, and safer media boundaries
• •Agent and Gateway recovery improved across 10+ failure scenarios including DM sends, media completions, session identity, and heartbeat deduplication
• •Provider/model handling expanded with GLM-5.2 and Claude Haiku 4.5 support, plus normalized provider-prefix handling across OpenRouter and Google Vertex
• •Usage hooks now include native full footer renderer, default templates, fixed-decimal formatting, and credential-aware limits with better error warnings
• •UI/mobile stability improved: workspace file collapsing, WebChat backscroll persistence, interactive sidebar session picker, and iOS foreground gateway reconnection
• •Memory and state diagnostics: oversized embedding batches split before errors, QMD memory search in transient mode, SQLite WAL disabled on NFS volumes
• •Discord auto-thread titles now have 60-second timeout and 4,096-token reasoning budget clamped to model output cap
• •Security enhancement: HTTP session/model override surfaces now require admin privileges
• •CLI backend improvements: `/btw` command support and proper CLI usage-error classification
• •Comprehensive documentation updates for node config, provider behavior, and channel/group functionality

Workflow Diagram

Concepts