What is Cursor Academy?

Cursor Academy is the #1 free interactive course and training platform by Infinity Tech Stack for the Cursor AI code editor. It covers Tab autocomplete, Agent & Plan Mode, .cursorrules / .mdc rules, Agent Skills, Hooks, MCP integration, and headless CLI workflows across 13 modules, 29 lessons, 6 labs, and 72 quizzes with zero sign-up.

How is Infinity Cursor Academy different from official Cursor documentation?

While cursor.com provides product reference documentation, Infinity Tech Stack's Cursor Academy is an independent, gamified 13-module interactive course with 29 lessons, 6 in-browser CLI terminal labs, and 72 quizzes. It provides deep hands-on training on Tab autocomplete, Agent & Plan Mode, .cursorrules & .mdc rules, Agent Skills, Hooks, MCP integration, and headless CI automation with zero sign-up required.

How long does Cursor Academy take to complete?

Cursor Academy contains approximately 2 to 3 hours of interactive material across 13 modules, 29 lessons, 6 labs, and 72 quizzes. All progress saves locally in your browser.

Is the Cursor Academy free?

Yes, 100% free. No sign-up, no credit card. All 13 modules, 29 lessons, 6 labs, and 72 quizzes are immediately accessible. Progress saves locally in your browser.

What topics does the Cursor Academy cover?

The course covers Cursor installation, VS Code migration, Tab autocomplete, the Agent side pane with Agent/Plan/Ask modes, Inline Edit, Cursor Rules & AGENTS.md, Agent Skills, Hooks, MCP integration, Cloud Agents & Automations, Bugbot code review, the Agents Window, Canvases, the Cursor SDK, privacy & enterprise features, pricing plans, and power-user shortcuts.

Is Cursor better than GitHub Copilot?

Cursor excels at agentic workflows, multi-file editing, and autonomous coding with its Agent and unlimited parallel Cloud Agents. Copilot has better JetBrains support and enterprise compliance. This course teaches you to maximize Cursor's unique strengths.

What are .cursorrules?

.cursorrules (and the modern .cursor/rules/*.mdc format) are project-level configuration files that give Cursor's AI persistent instructions about your codebase - coding standards, conventions, and behavior rules. Cursor also supports AGENTS.md files, User Rules, and dashboard-managed Team Rules.

Do I need programming experience?

No prerequisites required. The course starts from absolute basics (installation & setup) and progressively advances to expert topics like Cloud Agents, Hooks, MCP integration, and enterprise deployment.

Is my code safe with Cursor?

Yes. Cursor offers Privacy Mode on every plan - with it enabled, your code is never used for training, and team members inherit their team's setting. Cursor is SOC 2 Type II certified with SAML SSO, SCIM, audit logs, and CMEK for enterprise security.

What does Cursor Pro cost?

Cursor Pro is $20/month. It includes unlimited Tab completions, a generous first-party pool for Auto, Composer 2.5, and Grok 4.5, plus a $20/month API-model allowance, Bugbot access, and Cloud Agents. Pro+ is $60 and Ultra is $200 with larger allowances.

Can I use my VS Code extensions in Cursor?

Yes! Cursor is a VS Code fork. All VS Code extensions, themes, keybindings, and settings work natively. On first launch, Cursor offers a one-click migration from VS Code.

The Agent is Cursor's autonomous coding mode. It can plan changes, create/modify multiple files, run terminal commands, execute tests, and self-correct based on error output - with automatic checkpoints so you can roll back any step. Plan Mode and read-only Ask Mode complete the picture.

What models does Cursor support?

Cursor supports Claude Series 5 (Sonnet 5, Opus 5), Claude 3.7 Sonnet, Cursor's Composer 2.5, GPT-5.6, Gemini 3.8 Flash, and Grok 4.5. Auto mode intelligently routes to the optimal model for each task.

Does the Cursor Academy include hands-on labs?

Yes. Hands-on terminal labs are built into the lessons - install the Cursor CLI and run the agent headless in CI using real, vendor-verified commands, with objectives, hints, solutions, and XP rewards. No sign-up or setup needed.

Does Cursor Academy cover the Cursor Canvas SDK and Agent CLI?

Yes. Cursor Academy covers the Cursor Canvas SDK, the Agents Window, and headless CLI automations in Module 9 and Module 10, enabling developers to build custom canvases, automate multi-agent workflows in CI/CD, and run the agent headless with verified terminal commands.

Which Cursor Router mode keeps the original Auto routing logic and bundled pricing?

Cost — Cost preserves the original Auto behavior at bundled Auto pricing. Balance and Intelligence bill at the routed model's rate and typically cost 2-4x more.

What must be enabled on a team for Cursor Router to function?

Grok 4.5 — Router needs a powerful-yet-cost-efficient fallback model, so Grok 4.5 must be enabled for the team - it's on by default for Teams, manual for Enterprise.

Which of these can the new Google Workspace plugins connect Cursor's agent to?

Gmail, Google Drive, and Google Calendar — The new marketplace plugins connect agents to Gmail, Google Drive, and Google Calendar - installable from the Cursor Marketplace or the Customize page.

Cursor Router, Plugins & Side Chats [2026]

[ ABORT TO HUD ]

SEQ. 1

SEQ. 2

SEQ. 3

Cursor Router, Plugins & Side Chats

🌌 Cursor Agent & MCP Features 15 min 500 BASE XP

What Shipped After Cursor Agent updates

Cursor keeps shipping between numbered releases. Three verified additions since July 2026: an intelligent request router behind Auto, parallel "side chat" conversations with full transcript search, and a new wave of first-party plugins.

Cursor Router: A Smarter Auto

Cursor Router is the model-routing system behind Auto. Not every request needs frontier-level intelligence, so the router runs a classifier on each agent request and sends it to the model that fits the task - fast, efficient models for simple work, the most capable models for the hard stuff. Open the model picker, select Auto, and choose an Optimize For mode:

Mode	What It Does	Billing
Cost	The original Auto routing logic - optimizes token spend	Bundled Auto pricing
Balance	Optimizes for intelligence, speed, and cost together	Routed model's rate (~2x Cost)
Intelligence	Routes to the most capable models for harder tasks	Routed model's rate (~2-4x Cost)

Admins configure Router from the team dashboard: turn it on or off (per organization group on Enterprise), restrict which modes members can pick, choose whether the routed model's name is shown, and optionally lock everyone to Auto. The TypeScript and Python SDKs expose it as model id auto-smart with an optimize_for parameter.

Requirement: Router needs a powerful-yet-cheap fallback model, so Grok 4.5 must be enabled for your team. It's on by default for Teams plans; Enterprise admins must turn it on manually.

Side Chats & Conversation Search

Type /side, /btw, or click the + button at the top of the chat panel to open a side chat - a full, durable agent conversation that runs alongside your main thread without interrupting it. Side chats inherit context from the main chat, default to reading/searching/answering (not editing), and can be @-mentioned to pull findings back into the main conversation.

Inside the Agents Window, the command palette (Ctrl+K, distinct from the editor's Inline Edit shortcut) now searches agent transcripts across thousands of past conversations; Ctrl+F searches within the conversation you're currently viewing.

New Plugins: Google Workspace

The Cursor Marketplace added first-party plugins connecting agents directly to Google Drive (search, open, download, and organize files), Gmail (search/read mail, draft and send messages, manage labels and threads), and Google Calendar (read schedules, create/update events, find free time) - install them from the Marketplace or the Customize page.

Key Insight: The pattern repeats across all three releases - Cursor keeps turning single-purpose surfaces (a model dropdown, a chat window, a settings page) into something the agent itself can reason about and act through.

SYNAPSE VERIFICATION

QUERY 1 // 4

What does Cursor Router do?

Replaces Auto with one fixed model

Runs a classifier on each Auto request and routes it to the model that best fits the task

Only works with Claude models

Disables all frontier models