npx @buildinternet/releases get netlifyRecently added a deploy-status card for AI chat clients, an MCP feedback tool for agents, and GPT-6.1 Terra in AI Gateway.
Deploy status now shows up inside the conversation. Netlify MCP opens a visual card of up to five recent deploys in ChatGPT and other clients that support MCP Apps, with status, branch, creation time, error messages, and links to the live site or dashboard. Status filters narrow the list to successful, failed, or queued deploys.
Agents now have a way to report when they get stuck. Any agent connected through Netlify MCP, a connector, or a plugin can submit feedback or report issues it can't resolve, such as a missing capability or an unclear error.
AI Gateway kept adding models, all available without API keys across Function types.
Form notifications now require a team email address. Without an Enterprise or Legacy Business plan, new or edited email notifications must go to an address belonging to a Netlify team member. Existing notifications keep working until you edit them. Invite the address owner to your team, or switch the recipient to a member.
Next.js apps need 16.4.5 or 15.5.29 for seven November vulnerabilities, including cached data leaking between users and HttpOnly cookie exposure. One fix also needs @netlify/plugin-nextjs 5.16.6, which redeploys pick up automatically unless you pin it. The October batch is covered by 15.5.28/16.4.1.
Netlify now sends form notifications to email addresses outside the team on any pricing plan once the recipient confirms via a short verification email, and no notifications are delivered until they click Confirm. Notifications to team members are unchanged and need no confirmation.
Meta’s Muse Ember 1.0 model is now available through Netlify’s AI Gateway with zero configuration required. Use muse-ember-1.0 alongside the Muse Spark and Muse Wonder models already available in AI Gateway.
Netlify's updated Self-Serve Subscription Agreement, Privacy Statement, and Acceptable Use Policy take effect December 4, 2026. Once an organization verifies a domain, Netlify may show it that your account exists, its plan type, and the names, URLs, and last deploy dates of your public sites, and may move your account into the organization's account after notifying you.
GPT Image 2.6 Sunburst and GPT Image 2.6 Flare are now available through Netlify's AI Gateway with no configuration, API keys, or authentication setup. Both models also work in Scheduled, Background, and Edge Functions with Netlify's caching, rate limiting, and authentication.
Netlify now publishes a JSON schema for netlify.toml that describes supported properties and accepted values, so editors can suggest properties and flag invalid values before a deploy. The $schema key is ignored at build time, and some checks such as directory existence still only run during the build.
Netlify MCP renders a Deploys card in ChatGPT and other MCP Apps clients listing up to five recent deploys with status, branch, creation time, error messages, and status filters. The card refreshes automatically while deploys are in progress, stopping when they finish or after 15 minutes.
On plans without Enterprise or Legacy Business, form notification emails must now belong to a member of the Netlify team when a notification is created or edited. Existing notifications with outside addresses such as orders@company.com keep working, but such an address can no longer be saved on edit.
The Netlify MCP server includes a new feedback tool that lets agents submit feedback or report issues to Netlify when they hit a problem they can't resolve, such as an unclear error or a missing capability. It works wherever Netlify MCP is used, including connectors and plugins, with no extra setup.
Next.js 16.4.5 and 15.5.29 fix seven security issues including cached content leaking between users or root params, HttpOnly cookie exposure in Edge routes, cache poisoning of shared fallback shells, and cross-client response injection. Netlify recommends upgrading and redeploying, and apps using the pinned OpenNext adapter need @netlify/plugin-nextjs 5.16.6 to cover the Middleware case.
Agent Runners runs can now be powered by Meta's Muse family, from Muse Wonder 1.0 for complex software engineering to Muse Spark 1.3 and 1.2 for agentic coding workflows. Each model shows its estimated credit cost in the configuration dialog, and Auto lets the agent pick a model per run.
OpenAI's GPT-6.1 Terra is available through Netlify's AI Gateway and Agent Runners with no API key or auth setup, letting the OpenAI SDK be used directly in Netlify Functions. The model also works in Scheduled, Background, and Edge Functions with Netlify's caching, rate limiting, and authentication.
MiniMax M3.1 Flash is now available through Netlify's AI Gateway and Agent Runners with zero configuration, usable via the OpenRouter SDK in Netlify Functions without managing API keys or authentication. It also works across Scheduled, Background, and Edge Functions with Netlify's caching, rate limiting, and authentication infrastructure applied automatically.
The Netlify connector lets a Grok agent answer questions about projects, update supported project settings, and deploy to Netlify through conversation, working across Grok Bot, Grok CLI, and grok.com. Grok invokes it automatically when a task needs Netlify access, so users describe the task rather than naming the connector.
New public API endpoints let you start an agent run and send follow-up prompts, answer an agent's questions, and review the files it changed. Runs start on the production branch by default, and a deploy URL in the response previews each run's work.
TypeSafe's Jev 1.14.0 model is available through Netlify's AI Gateway with zero configuration, billed to Netlify credits, and usable via @typesafe-ai/sdk in Netlify Functions without managing API keys or authentication. It is also available across Scheduled Functions, Background Functions, and Edge Functions.
Enterprise teams can now create, track, and reply to support tickets without leaving Netlify, under User settings > Support, with statuses of Open, Awaiting customer response, and Closed, plus search, sort, filtering, and CC recipients. Team Owners can see every ticket their team members created, while members only see their own.
Three Next.js and React vulnerabilities disclosed October 14, 2026 — two critical and one high, covering unauthenticated RCE in use cache functions returning typed arrays, RCE via SVG processing reachable through next/og ImageResponse, and DoS in App Router Server Actions — are patched in 15.5.28 and 16.4.1. Netlify recommends upgrading next and redeploying, and notes that public deploy previews and branch deploys may remain vulnerable until auto-deleted.
Mistral Large 4 is now selectable from the model picker in Netlify Agent Runners, with a 524K-token context window aimed at agentic coding and long-context, multi-step tasks across larger codebases.
TanStack Start apps now deploy on Netlify, with SSR, server routes, server functions, and middleware running on Netlify Functions. A @netlify/vite-plugin-tanstack-start plugin emulates Netlify primitives in the dev server, and new Agent Runners projects default to TanStack Start 1.0.
Google's Gemini 4 Argon model is now available through Netlify's AI Gateway and Agent Runners with zero configuration, so the Google GenAI SDK can be used directly in Netlify Functions without managing API keys or authentication. It is available for all Function types and Agent Runners, with Netlify's caching, rate limiting, and authentication handled automatically.