npx @buildinternet/releases get cursorCloud agents can now subscribe to PRs, Slack threads, or scheduled tasks, wake on events, and drive work to completion. Skills can be pinned as custom modes, subagents can run on isolated VMs, and /goal gives agents long-lived objectives; follow-up messages now wait for the next tool call instead of interrupting.
Cursor now hosts code in early beta on paid plans, starting with repos, pull requests, code browsing, and GitHub sync. GitHub repos can sit alongside Cursor-hosted ones with two-way PR comment sync, and agents can answer questions about code, make changes, and push branches. Vercel, Depot, and Buildkite integrations are available now.
Cloud Agents now boot from Builds — ready-to-use environment copies Cursor prepares in the background — cutting time to first token by 3x. When a bad commit or dependency update breaks the environment, agents keep using the last successful build; a Builds tab in the dashboard shows status, logs, and commit SHAs.
Cursor can now read, write, and act across Google Workspace via plugins for Google Drive, Gmail, Calendar, Docs, Sheets, and Chat — search, create, edit, and manage content without leaving Cursor.
Cursor for iPad is available on all paid plans, with a rebuilt layout for the larger screen including pinned sidebar chats, split-screen review, and full file diffs. Both iPhone and iPad gain an inbox and a complete PR review surface covering comments, checks, and approvals.
Cursor introduces a new Start plan for developers in India at ₹649 per month with local pricing and UPI. Includes Grok 4.5, cloud agents, iOS access, and plugins.
Auto mode is now powered by Cursor Router, which analyzes each request and routes it to the appropriate model. Three optimization modes (Intelligence, Balance, Cost) control cost-intelligence tradeoffs, and admin controls allow per-team enablement, mode restrictions, and model allow/block lists.
Cursor in Slack now shares a plan before starting work, runs in multi-repo environments, and can work across channels and threads. In-message buttons replaced by compact footer links.
Side chats let you ask questions and explore tangents without interrupting your main agent conversation, each a durable full agent session. Conversation search indexes agent transcripts across thousands of conversations via the command palette, plus redesigned project and repo pickers with scoped search and simplified workflows. New cloud agent hooks for observing prompts, responses, thinking, and subagents.
Admins can now configure Team MCP servers once and distribute them across cloud agents, the agents window, IDE, and CLI through team marketplaces. Marketplace access can now also be restricted to organization groups in addition to SCIM directory groups.
Cursor is now available as an iOS app for paid plans, letting users launch and manage cloud agents from their phone. Includes Remote Control to direct a desktop agent and Live Activities on the lock screen.
The new Customize page unifies management of plugins, skills, MCPs, subagents, rules, commands, and hooks at user, team, or workspace levels. Adds a marketplace leaderboard for popular add-ons, prebuilt plugin canvases (Hex for data viz, Atlassian for realtime issues), and GitLab/BitBucket/Azure DevOps support for team marketplace imports.
Cursor Automations now include the /automate skill for creating automations in-agent via plain language, Slack emoji triggers, five new GitHub triggers, and a computer use tool for producing demos and artifacts. Plus, automations can be saved in an incomplete state and open PRs by default.
Cursor can now set up cloud dev environments in under 10 minutes with reusable snapshots, and cloud subagents via /in-cloud run in their own VM and branch. Subagents can babysit PRs remotely, and agent sessions can handoff between local and cloud machines.
Bugbot is now over 3x faster with an average review time of 90 seconds, finds 10% more bugs per review (0.62 on average), and costs ~22% less per run, thanks to progress in training Composer 2.5. You can now run Bugbot with /review before pushing code, and configure it to only review what's new in your PR.
Design Mode now supports multi-selecting two or more elements in the browser to adjust groups of components at once, and voice input to narrate changes through the overlay while agents are running.
Design Mode is now available in canvases, allowing you to select and annotate UI elements directly to guide Cursor's edits. Cursor can now show your agent's context usage as an interactive report in a canvas, breaking down where tokens go across the system prompt, tool definitions, rules, and skills.
The Cursor SDK now supports custom tools passed as function definitions through local.customTools, auto-review routing for local tool calls via a classifier steered by permissions.json, and nested subagents that can spawn their own subagents. Both SDKs also gain JSONL and custom store support for agent metadata persistence, plus reliability fixes for run correlation, safe checkpoints on dispose, HTTP/1.1 cloud streaming, and lighter imports for cloud-only consumers.
Enterprise customers can now manage multiple Cursor teams under a single organization, with separate security, governance, budget, and feature controls per team. Organizations provide unified spend and token usage rollup, IDP management at the organization level, and per-team or per-group model access and spend limits. Users can belong to multiple teams with different roles in each, and admins can move users between teams via dashboard, API, or CSV.
Auto-review is a new run mode that applies to Shell, MCP, and Fetch tool calls. Allowlisted calls run immediately, sandboxed calls execute in isolation, and remaining actions go to a classifier subagent that decides whether to allow, retry, or ask for approval. Configure it in Settings > Cursor Settings > Agents > Run Mode.