• Google restricting AI Pro/Ultra users for OpenClaw usage [TRENDING] — Multiple subscribers report account restrictions without warning after using OAuth with OpenClaw, generating massive community concern with 2212+ engagement. Discussion thread
• Security researchers discover malicious OpenClaw skills in marketplace [TRENDING] — Bitdefender Labs exposes hundreds of malicious skills disguised as helpful utilities, with community debating whether OpenClaw is "dangerous" (269+ engagement). Analysis
• OpenClaw transforms user productivity workflows [TRENDING] — Community sharing real-world success stories of life-changing automation, with 1366+ engagement on testimonials about 24/7 AI assistance. Personal experience
• 189-page OpenClaw guide with 23 monetization cases goes viral [TRENDING] — Chinese community sharing comprehensive blueprint covering installation, model configuration, and industry applications with 432+ social engagement. Guide announcement

• OpenClaw Blue Book [TRENDING] - 189-page comprehensive guide with 23 real money-making cases, installation guides, and 10 industry solutions gaining massive traction. 216 retweets View tweet
• Model Performance Secrets [TRENDING] - Developer shares insider knowledge about AI models including "Opus 4.5 > Opus 4.6" and "Qwen 3.5 is best OS model series." 122 retweets View tweet
• Hardware vs Software Debate [TRENDING] - User advocates canceling ChatGPT subscriptions and "deleting OpenClaw slop" in favor of buying used RTX 3060 hardware. 58 retweets View tweet
• FlashClaw vs Perplexity [HOT] - Claims FlashClaw delivers outcomes without installations or configuration screens, unlike traditional tools. 37 retweets View tweet
• Financial Analysis Integration [HOT] - New Financial Datasets MCP Server lets Claude/Cursor/OpenClaw directly query stock prices, earnings, and SEC filings. 19 retweets View tweet

• openclaw v2026.3.13-1 - Recovery release fixes broken v2026.3.13 due to GitHub's immutable release constraints Download
• openclaw v2026.3.12 - Major Control UI/dashboard-v2 refresh with modular overview, chat, config, agent and session views, plus command palette and mobile bottom tabs Download

Project Stats: 313.7k stars (+143k growth), 59.9k forks, 1,257 contributors

• yfge fixes Feishu interactive cards — Webhook parser now surfaces card content instead of raw JSON, enabling card-based workflows. PR #41678
• nnemirovsky adds MCP tool approval system — New security gating for MCP/plugin calls with configurable policies, matching existing exec approval system. PR #46970
• Br1an67 adds per-agent media isolation — Inbound media files now saved to agent-specific directories by account ID, preventing cross-agent access in multi-agent setups. PR #46162
• nnemirovsky enables SecretRef for MCP servers — MCP server environment variables now support SecretRef providers (env/file/exec) for secure API key management across 73+ available MCP servers. PR #46825
• bingolam adds MiniMax TTS provider — Fourth TTS option alongside OpenAI, ElevenLabs, and Edge TTS, expanding voice synthesis capabilities. PR #44716

• Google restricting Google AI Pro/Ultra subscribers for using OpenClaw — Users report account restrictions without warning when using OpenClaw with Google AI services, sparking debate about API usage policies (800+ HN points, 700+ comments)
• OpenClaw is changing my life — Personal story of productivity transformation through AI automation, resonating strongly with the community (340+ HN points, 500+ comments)
• OpenClaw is dangerous — Critical analysis of potential risks and safety concerns with autonomous AI agents (83+ HN points, active discussion)
• Medina-Qwen3.5-27B-OpenClaw — New fine-tuned model optimized for OpenClaw tool calling and function execution (330+ downloads on HuggingFace)
• OpenClaw on DigitalOcean — Official marketplace deployment option for one-click OpenClaw hosting on managed infrastructure

• OpenClaw Malicious Skill Trap - Hundreds of malicious skills are disguising themselves as legitimate OpenClaw extensions, making manual review impractical. Review your installed skills immediately and enable skill verification.