2 min lesson
The demo walkthrough: a runnable spine
Give a plain answer to "Why open your demo in Plan mode rather than letting Agent execute immediately, and what's the one moment you most want to provoke?" Then ground it in one lesson detail.
Step 1 of 2
A great demo is a spine, not a script: plan first, let the model surface a discrepancy, verify in the browser and hand back proof. The four beats below survive an unexpected codebase and let you narrate judgment at every step.
The single biggest workshop recommendation across Cursor's own field sessions is the same one that makes a demo land: start in Plan modeA mode that makes no edits: it researches the codebase and produces an editable plan you review before any code changes. Press Enter for the full definition.. It's the antidote to the "agent went rogue across six files" story, and it's where you can show thinking out loud before any code is written.
- 1Beat 1 - Plan modeA mode that makes no edits: it researches the codebase and produces an editable plan you review before any code changes. Press Enter for the full definition.. Switch to Plan mode and give the goal. It makes no edits: it researches the whole indexed codebase, follows existing paradigms and outputs a hyperlinked Markdown plan (scope, implementation, tests, concrete to-dos, often a data-flow diagram). Editing the plan costs zero tokens, so you converse or edit the Markdown directly - and teammates can sign off before execution.
- 2Beat 2 - Surface a discrepancy. Let Plan modeA mode that makes no edits: it researches the codebase and produces an editable plan you review before any code changes. Press Enter for the full definition. ask its clarifying questions instead of rushing past them. This is the moment that wins skeptics.
- 3Beat 3 - Execute, then verify in the browser. Switch to Agent to run the to-dos, then open the built-in browser, give a plain-language directive and watch the change render while Cursor surgically edits only the necessary files.
- 4Beat 4 - Hand back proof. Show artifacts: the agent self-tests, records a video and screenshots and can upload the walkthrough straight to the PR.
In a real session, a PRDProduct Requirements Document. The spec describing what to build, for whom and why, before implementation starts. Increasingly the artifact you hand an AI agent as the source of truth for a feature. Press Enter for the full definition. said to "use authenticated users," but the agent saw no authentication in the app - so it paused and asked how to handle the gap (build full auth now, or ship a local-first V1?). Reframe clarifying questions as a feature, not friction: the gap-catching is a key reason planning works.
Tee it up by handing Plan modeA mode that makes no edits: it researches the codebase and produces an editable plan you review before any code changes. Press Enter for the full definition. a deliberately under-specified prompt, then narrate: "Notice it didn't guess - it found the contradiction and asked. That's the judgment we're rolling out, not a tool that fills in blanks silently."
Learn more
Full explanation
Why plan-first is also token-efficient
Generating the plan with a frontier/thinking modelA reasoning model (shown with a brain icon in Cursor's picker) that spends extra compute before answering; reach for it on complex, nuanced work and a standard model for fast, simple tasks. Press Enter for the full definition. is cheap because the expensive model only reasons about the approach - it doesn't output all the code. Then a faster, cheaper coding model executes the to-dos. "Since we did the plan, we were token efficient; we did not have to output all the code, just reason about it."
Credibility support to cite: a University of Chicago study found more experienced developers are more likely to plan before generating, especially with AI tools - so plan-first isn't a beginner crutch, it's what strong engineers already do.