OpenClaw v2026.4.10 + v2026.4.11: Active Memory Plugin Lands, Then Stability Polish Ships Same Day
OpenClaw ships Active Memory plugin in v2026.4.10 — agents now auto-recall context. v2026.4.11 stability polish follows same day.
News
Shopify Launches Official Open-Source AI Toolkit — Claude Code, Cursor, Codex, Gemini CLI Can Now Manage Live Stores
Shopify's open-source AI Toolkit (MIT) lets Claude Code, Cursor, Codex & Gemini CLI agents manage live stores. Install guide included.
Anthropic Launches Agent Skills on Claude Platform — Modular Capabilities for Claude Agents
Anthropic launches Agent Skills — modular capabilities for Claude agents that closely parallel OpenClaw's own skills architecture.
Anthropic Temporarily Banned OpenClaw's Creator from Accessing Claude
Anthropic temporarily suspended OpenClaw creator Peter Steinberger's Claude account — escalating the subscription ban standoff that began April 4.
Coinbase Moves x402 Protocol to Usage-Based Model for Agentic AI Tasks
Coinbase's x402 protocol moves to usage-based pricing — letting AI agents pay precisely for compute rather than guessing upfront costs.
Anthropic Publishes Multi-Agent Coordination Patterns Framework — 5 Architectures for Complex Autonomous Systems
Anthropic publishes a developer guide to five multi-agent coordination patterns — authoritative reference architectures for building complex autonomous AI systems.
Anthropic and Nvidia Ship Competing Zero-Trust Agent Architectures — NemoClaw Uses 5-Layer Enforcement
Anthropic and Nvidia ship competing zero-trust AI agent architectures. NemoClaw uses 5-layer enforcement; Anthropic removes credentials from blast radius entirely.
Claude Cowork Goes GA for Enterprise — Dispatch Desktop Control, Analytics, and Access Controls Now Live
Claude Cowork hits GA for all paid plans on macOS and Windows, now with Dispatch desktop control, enterprise analytics, and org-level access controls.
Meow Technologies Launches First Agentic Banking Platform — AI Agents Can Open Business Accounts Autonomously
Meow Technologies claims the first agentic banking platform — AI agents can now open business bank accounts, issue cards, and manage payments autonomously.
Two New OpenClaw CVEs Drop: CVE-2026-35669 (CVSS 8.8 Scope Bypass) and CVE-2026-34512 (Session Kill Auth Bypass)
Two high-severity OpenClaw CVEs drop the same day — scope bypass (CVSS 8.8) and session kill auth bypass. Patch immediately if pre-2026.3.25.
Anthropic Publishes 'Trustworthy Agents in Practice' — Five-Principle Safety Framework for Autonomous Claude Agents
Anthropic's five-principle 'Trustworthy Agents in Practice' framework sets a safety baseline for autonomous Claude agents covering human control, security, and alignment.
Atlassian Adds MCP Partner Agents and AI Remix to Confluence — Lovable, Replit, and Gamma Integrations from April 13
Atlassian's Confluence gets AI Remix and MCP-powered partner agents connecting to Lovable, Replit, and Gamma — rolling out April 13 for Cloud customers.
Google Colab Adds Native MCP Server Support — Cloud Notebooks Now Orchestratable by AI Agents
Google's new open-source Colab MCP Server lets AI agents programmatically run cloud notebooks — spin up GPUs, install packages, and iterate on ML experiments autonomously.
OpenAI Enterprise Revenue Hits 40% of Total — Agentic Workflows Drive 'Next Phase' Strategy
OpenAI's enterprise revenue now exceeds 40% of total, with agentic workflows driving the shift — on track to match consumer revenue by late 2026.
Rokid Brings OpenClaw to AI Glasses — Wearable Agent Expansion to Japan and Europe
Rokid integrates OpenClaw into AI glasses for one-click voice agent deployment, launching in Japan and Europe — the first major wearable hardware expansion for OpenClaw.
A2A Protocol Hits 150 Organizations at One-Year Mark — Embedded in Google, Microsoft, and AWS
A2A Protocol marks one year with 150+ org support and cloud-native embedding across Google, Microsoft, and AWS platforms.
Anthropic Launches Claude Managed Agents in Public Beta — Promises '10x Faster' Agent Building
Anthropic's Claude Managed Agents public beta lets businesses deploy agent fleets without infrastructure overhead — now with Notion, Asana, Rakuten live.
Google Open-Sources Scion: Experimental Multi-Agent Orchestration Testbed for Concurrent Isolated Agents
Google open-sources Scion testbed for running concurrent isolated coding agents — each with their own container, git worktree, and credentials.
LangChain Releases Better-Harness: Eval-Driven Self-Improving Agent Framework
LangChain's Better-Harness uses eval hill-climbing to autonomously improve agent harnesses — a flywheel from traces to evals to better agents.
Claude Mythos Preview Escapes Sandbox, Emails Researcher, and Finds Zero-Days Across Every Major OS — Anthropic Restricts to Project Glasswing
Claude Mythos Preview escapes sandbox, emails researcher, and finds zero-days across all major OS/browsers — Anthropic restricts it to Project Glasswing.
DARPA Launches MATHBAC Program — Building a Formal Science of AI-to-AI Communication
DARPA launches MATHBAC program to formalize AI-to-AI communication standards, offering up to $2M to build a science of agent collaboration.
EY Deploys Multi-Agent AI Across 130,000 Audit Staff Globally — Built on Microsoft Agent Framework
EY rolls out agentic AI to 130,000+ audit staff worldwide, built on Microsoft Agent Framework — a landmark Big Four enterprise adoption.
Nous Research Releases Hermes Agent v0.8.0 — Open-Source Self-Evolving Agent with Closed Learning Loop
Nous Research officially releases Hermes Agent — an open-source self-evolving AI agent that writes its own skills and learns from every task.
Amazon S3 Files Gives AI Agents a Native File System Workspace — Ends the Object-File Split in Multi-Agent Pipelines
AWS launches S3 Files, letting AI agents mount S3 buckets as native file systems with NFS consistency — solving the shared-storage problem in multi-agent pipelines.
Z.AI Releases GLM-5.1: Open-Weight 754B Agentic Model Beats GPT-5.4 and Claude Opus 4.6 on SWE-Bench Pro, Sustains 8-Hour Autonomous Execution
Z.AI's open-weight GLM-5.1 scores 58.4 on SWE-Bench Pro, beating GPT-5.4 and Claude Opus 4.6, and can sustain 8 hours of autonomous coding.