JetBrains opened Early Access for Air in its IDEs (Marketplace plugin or 2026.3 EAP): an agentic-first shell that detects Codex, GitHub Copilot, Cursor, Claude Agent, Junie, and other ACP agents, with parallel sessions, temporary worktrees, and free Junie Lite runs. It ships with no agents installed and sends data to your existing provider, not JetBrains.
DeepKeep launched AI Lens for Developers: a light plugin that hooks Cursor and Claude Code to flag credential leaks in prompts/files, catch insecure generated code, and send destructive shell/MCP calls for human approval before they run. Copilot, Codex, Lovable, and Windsurf support is listed as coming soon…
Qodo 3.0’s PR Triage (research preview) groups related multi repo agent PRs into work packages with blast radius and review order, and can hand a brief to Claude Code or Codex. An Agentic Toolbox also pulls org rules into the coding agent so issues get reviewed before a PR opens…
Google’s Cloud CLI remote MCP server is in public preview at cloudcli.googleapis.com/mcp, exposing run gcloud command and run bq command so agents can run hundreds of admin commands without a local gcloud install. Enable the Cloud CLI Execution API, grant MCP Tool User, and commands run under the caller’s IAM in a network isolated sandbox…
Barclays is expanding its Anthropic partnership so Claude Code reaches 50% of developers by end of 2026 and a majority in 2027, alongside Claude for legacy modernization and ops. The bank already runs a Claude knowledge assistant for 16,000+ UK colleagues and routes 120,000 Global Markets emails a day…
Transluce reports AI agents made rudimentary probes against Library and Archives Canada’s collection search on May 28 and June 9 (899 requests, including SQL injection); none succeeded and Canada says no systems were compromised. Tactics resembled prior OpenAI linked agent activity; OpenAI says it is reviewing the findings and briefed Canadian officials…
Spotted something we missed? Start a thread.