Frequently Asked Questions

When AI tools like Cursor or Claude generate wrong code, it's almost always because they don't have enough context about your product. They're guessing based on patterns from their training data, not your actual requirements. SuperPlan fixes this by helping you build a structured product index that captures your specific constraints, edge cases, and business logic. Once your AI tools have this context, they stop guessing and start building what you actually need.

Think of a product index as a single source of truth about your product that both you and AI tools can understand. It includes your core thesis, who you're building for, key features, technical constraints, and important edge cases. Without one, you're constantly re-explaining your product to AI tools, and they keep making the same mistakes. With a product index, your AI agents have the context they need to generate accurate code from day one.

Notion and Linear are great for team collaboration, but they weren't built for AI consumption. Your AI tools can't effectively use scattered notes across different docs. SuperPlan takes all that messy product thinking and converts it into structured context that AI agents can actually query. You keep using your existing tools for collaboration, SuperPlan just makes that work usable by Cursor, Claude, and other AI coding tools.

Not at all. ChatGPT wrappers give you a chat interface that forgets everything after each conversation. SuperPlan creates a persistent product index that lives across all your AI tools and sessions. You build it once, and every AI tool you use benefits from the same structured context. It's more like building a knowledge base for your AI agents than having another chat interface.

AI tools invent things when they don't have enough information to work with. If you ask Cursor to build a feature without explaining your existing architecture, it'll make assumptions based on common patterns it's seen before. Sometimes those assumptions are wrong. SuperPlan solves this by giving your AI explicit context about what exists in your product, what doesn't, and what the boundaries are. When the AI knows the rules, it stops making things up.

It depends on how much context you provide, but developers using structured product context typically report spending 40 to 60 percent less time fixing AI generated code. The key is being explicit about constraints and edge cases. When your AI knows that certain APIs don't exist or that specific user flows have particular requirements, it generates code that actually fits your product instead of generic solutions that need heavy modification.

We aim for practical PM quality, not polished marketing copy. The output clearly states assumptions, lists constraints, defines edge cases, and gives AI tools exactly what they need to work accurately. It won't win any writing awards, but it will make your AI agents significantly more useful. Most importantly, the docs are structured in a way that AI tools can actually parse and use effectively.

Most people have a working product index within 15 to 30 minutes. You'll answer a series of guided questions about your product, who it's for, what you're building, and what constraints matter. SuperPlan structures your answers into a format your AI tools can use. After that initial setup, you can export to Cursor, Claude, or other tools right away and start seeing better results immediately.

You'll spend a bit more time upfront, yes. But that time pays off quickly. Instead of spending hours debugging AI generated code that missed important requirements, or re-prompting the same context over and over, you do it once and your AI tools just work better. Teams using structured context consistently report that the upfront investment saves them significant time on every feature they build afterward.

Yes. SuperPlan accepts all kinds of messy inputs, including existing PRDs, meeting recordings, voice memos, competitor research, and scattered notes. You can type directly, upload files, or connect integrations. The system extracts the relevant product context regardless of how disorganized your starting point is. Most founders have product thinking scattered across dozens of places, and that's fine.

Yes, and that's one of the biggest benefits. Your product index isn't locked to any single tool. Export it once and use it in Cursor for coding, Claude for brainstorming, Windsurf for quick edits, whatever fits your workflow that day. Every tool gets the same context, so you're not re-explaining your product when you switch between them. Most developers use two or three AI tools regularly, and SuperPlan keeps them all on the same page.

Export your product index from SuperPlan as a context file and add it to your Cursor project. You can put it in your .cursorrules file or reference it directly in your prompts. Cursor will use this context when generating code, which means it'll understand your product's specific requirements instead of guessing. The difference is noticeable immediately, especially for complex features.

We're building an MCP server first, which will let any MCP compatible tool query your product index directly. That covers Claude Desktop, Cursor, and a growing list of AI tools. API access for custom integrations is planned after that for teams who want to build SuperPlan into their own workflows. If you have specific integration needs, reach out and let us know what you're building.

Absolutely. Your data is yours. SuperPlan supports exporting in multiple formats including Markdown for general use, JSON for programmatic access, and optimized formats for specific AI tools. You can export everything at any time. We're also building one click export to popular AI coding environments to make the workflow even smoother.

MCP stands for Model Context Protocol, and it's an open standard created by Anthropic that lets AI tools connect to external data sources. Instead of manually copying context into every AI conversation, MCP lets tools like Claude and Cursor automatically pull the information they need. For SuperPlan users, this means your AI agents can access your product index on demand without you doing anything. It's a game changer for workflow efficiency.

Once our MCP server launches, you'll add it to your Claude Desktop config file and that's it. After that, when you ask Claude questions about your product, it can automatically pull relevant context from your SuperPlan index. No more pasting requirements into every conversation. Just ask Claude about your features, constraints, or edge cases and it knows what you're talking about.

Yes. Cursor supports MCP servers, so the same integration that works with Claude Desktop will work with Cursor. Once configured, Cursor's AI can query your product index while it's generating code. This means better code on the first try because Cursor actually understands your product context, constraints, and edge cases without you having to explain them every time.

Use export now and switch to MCP when it's available. Export works great, you just need to manually update your context files when your product changes. MCP automates that so your AI tools always have the latest context. Both approaches give you better AI output than working without structured context. MCP just removes the manual step of keeping everything in sync.

We're actively building it and plan to release it for early access users soon. Join the waitlist at superplan.md to get notified when it launches. In the meantime, the export feature works great for getting structured context into your AI tools. Many users find that even manual export dramatically improves their AI workflow.

Think of it this way: you're already paying for powerful AI tools, but they're only as good as the context you give them. SuperPlan makes those existing subscriptions more valuable by helping your AI actually understand what you're building. If you're spending even an hour a week fixing AI generated code that missed requirements, SuperPlan pays for itself quickly. It's not a replacement for your AI tools, it's what makes them finally work the way you expected.

Yes, we're planning a free tier that lets you create a basic product index and export to one AI tool. This way you can experience the value before committing to a paid plan. The paid tiers unlock unlimited exports, more AI processing capacity, team collaboration features, and advanced integrations.

Team plans are coming soon with features like shared product indices, collaborative editing, and role based access control. For larger companies, we'll offer enterprise features including SSO, audit logs, custom integrations, and dedicated support. If you're interested in early access to team features, drop us a note at [email protected].

We take security seriously. All data is encrypted both in transit and at rest. We use strict access controls and run regular security reviews. Your product data lives on AWS and Google Cloud with enterprise grade security. We never share your data with other customers, and we don't use it to train AI models unless you explicitly opt in.

We use OpenAI, Anthropic, and AssemblyAI for different AI processing tasks. These providers process your inputs to generate the structured outputs. Their respective privacy policies govern how they handle data during processing. If you have specific concerns about third party processing, email [email protected] and we can discuss options for your workspace.

Yes. You can delete individual projects or your entire account whenever you want. When you delete data, it's removed from our active systems right away. Some data may stay in encrypted backups for a limited period as required for legal compliance, but it's not accessible or used. For formal deletion requests, contact [email protected].

Projects with real complexity benefit most. If you're building an MVP with specific business logic, a product with multiple user types and different flows, or anything where AI keeps missing important details, SuperPlan will help a lot. Simple scripts or basic CRUD apps probably don't need it. But if you've ever spent hours fixing AI generated code because it didn't understand your requirements, SuperPlan is built for you.

SuperPlan works great for existing products too. You can import your current documentation, feed in context from codebase discussions, and build a product index that captures all the institutional knowledge that currently lives only in people's heads. A lot of teams use SuperPlan to finally document what they've already built so that AI agents can work on it without breaking things.

Definitely, SuperPlan is designed to capture product thinking, not technical implementation details. You describe what you're building, who it's for, and what problems you're solving. The structured output helps whether you're coding yourself with AI assistance, working with AI agents directly, or handing off to developers. Clear product context makes everyone more effective, technical or not.