AI Coding Live News

Cline moved its open coding agent into the browser: connect GitHub, pick a repo and model, and each task runs in an isolated cloud sandbox that commits to a cline/ branch and can open a PR—even with your laptop closed. Limits: 10 concurrent sessions and 50 new sessions/day. Providers: ClineFree, ClinePass ($9.99), and usage-billing. Cloud sessions do not yet load Rules, Skills, MCP, or Connectors.

@DanKornas highlighted yetone/cumora, cross-platform team chat (Electron, PWA, iOS, Android) where humans and AI agents share the same roster, DMs, groups, Kanban, and calendar. Run managed Cumora Cloud agents or pair a Mac/VPS with npx cumora agent computer for BYOA brains (Claude Code, Codex, Cursor Agent, OpenCode, and more) without shipping provider keys. Coordination includes work claims and a freshness gate so agents do not trample each other. MIT-licensed with about 3764 GitHub stars at verification.

OpenAI launched the Agents API—a managed cloud agent service powered by the Codex harness for orchestration, long-running sessions, and tool use. Bring connectors, attach any sandbox, and run agents with Astra.

Gemini 3.8 Flash is now in the GitHub Copilot model picker across Pro, Pro+, Max, Business, and Enterprise—on VS Code, Visual Studio, Copilot CLI, cloud agent, Copilot app, JetBrains, Xcode, and Eclipse. Unlike GPT-6 Astra, Flash is available on plain Pro; introductory usage-based provider pricing runs through 2026-12-31.

Capy cofounder Justin Sun (@justinsunyt) highlighted a new feature that lets cloud coding agents access your local machine, bridging cloud agents and the developer workstation.

Cua (@trycua) announced Cursor Cloud Agent workers on Windows and Linux Cloud Fleets: Cursor runs the agent loop while tools execute inside a Cua VM. Requires Cursor Enterprise + Self-Hosted Agents, a service-account API key, and Fleet credentials.

Cursor now lets you run cloud agents on your own infrastructure, including pools of machines that automatically scale with demand. Agents can access internal services or specialized hardware while the agent loop stays in Cursor.

Cursor announced that its cloud coding agents can run on customer-managed machine pools that autoscale with demand. The agent loop stays in Cursor while workers gain access to internal services or specialized hardware, with support for sandboxes including AWS Lambda, Cloudflare, E2B, Modal, and Vercel.

Cursor now lets cloud agents run on customer-managed machine pools that autoscale with demand, while the agent loop stays inside Cursor. Teams can grant agents access to internal services or specialized hardware, and can also use sandbox providers including AWS Lambda, Cloudflare, Daytona, E2B, Modal, and Vercel. It is a major step from fully hosted coding agents toward enterprise-controlled runtimes.

Cursor (@cursor_ai) announces cloud agents can now run on your infrastructure, including pools of machines that automatically scale with demand. Agents can reach internal services or specialized hardware while the agent loop stays in Cursor.

Elon Musk said Grok 4.7 comes out in 10 days after Shopify CEO Tobi Lütke publicly praised Grok 4.6. In parallel, xAI’s Grok Bot now runs 24/7 in the cloud and can act across Microsoft Outlook, Calendar, and OneDrive — a fast-moving agent stack some investors are calling another ChatGPT moment.

Elon Musk said Grok 4.7 will ship in 10 days, responding to Shopify CEO Tobi Lütke's praise of Grok 4.6. The teaser lands as xAI's Grok Bot—always-on cloud agents with Microsoft Outlook, Calendar, and OneDrive plugins—is being called another ChatGPT moment for turning hours of coding-agent work into seconds.