What is new in OpenClaw 3.22?

OpenClaw 3.22 (March 23, 2026) adds Matrix messaging via matrix-js-sdk, Anthropic Vertex AI support for running Claude through Google Cloud, bundled web-search providers (Exa, Tavily, Firecrawl), and native ClawHub CLI skill install. It includes 45 new features, 82 bug fixes, and 13 breaking changes.

Does OpenClaw 3.22 support Matrix messaging?

Yes. OpenClaw 3.22 adds a native Matrix messaging plugin backed by the official matrix-js-sdk, enabling communication through the decentralized Matrix protocol for privacy-focused operators and self-hosted infrastructure teams.

How do I install OpenClaw skills from ClawHub in 3.22?

With OpenClaw 3.22, you can install skills directly from the command line using the native ClawHub CLI integration. This eliminates the need to manually download and configure skills from the ClawHub registry.

Does OpenClaw 3.22 support Google Vertex AI?

Yes. OpenClaw 3.22 adds Anthropic Vertex AI support, allowing enterprises to route Claude requests through Google Cloud instead of Anthropic's direct API. This is useful for organizations that require all AI traffic to go through Google Cloud for compliance and billing.

What breaking changes are in OpenClaw 3.22?

OpenClaw 3.22 removes all CLAWDBOT_* and MOLTBOT_* prefixed environment variables. If your .env file uses these legacy variables, your agent will fail to start after upgrading. You must rename all environment variables to use the current OpenClaw naming.

← Back to dashboard

clawsmith.com/signal/openclaw-3-22-matrix-vertex-ai-clawhub

📈 TrendsWide OpenLive

OpenClaw 3.22: Matrix Messaging, Vertex AI Support, Native ClawHub CLI — 294 GitHub Reactions

OpenClaw 3.22 is the most significant update since the MoltBot rename. Complete plugin SDK overhaul (old extension API abandoned), native ClawHub CLI integration, cross-tool plugin discovery for Cursor/Claude/Codex plugins, bundled web-search (Exa/Tavily/Firecrawl), pluggable SSH-based sandbox backends. 294 GitHub reactions.

Product Idea from this Signal

A security layer that vets ClawHub skills for malware and prompt injection before your agent installs them

79.8k ▲

ClawHub grew 380% to 13,729 skills in Q1 2026. Snyk found 36% contain prompt injection and 1,467 carry malicious payloads. The ClawHavoc campaign planted 1,184 weaponized skills in the marketplace. VirusTotal integration catches known malware but misses novel prompt injection, data exfiltration via tool outputs, and social engineering patterns unique to AI agent skills. This tool performs deep behavioral analysis of every skill before installation, catching threats that signature-based scanners miss.

SECURITYCLIDEVTOOLOPEN-SOURCE

CompetitiveView Opportunity →