What Claude does
When you submit an idea, our server sends it to the Claude API along with a fixed set of instructions describing how a pragmatic product strategist scopes an MVP. Claude reasons through the trade-offs and returns a plan in a strict, structured format. Our app checks the structure and shows it to you.
- Model: Claude Opus 5.5 via Anthropic's first-party Claude API.
- Structured output: every plan must match a fixed schema, so you never get half a plan dressed up as a whole one.
- Your text is data: Claude is instructed to treat your idea as a description, not as instructions to follow.
What Claude doesn't do
- It doesn't see anything except the idea, audience, skill level and budget you submit.
- It doesn't browse the web or verify market facts — treat numbers in a plan as estimates.
- It doesn't make decisions for you. A plan is a starting point for your own judgement.
Safeguards
- Use of ShipMVP must follow our Acceptable Use Policy, which incorporates Anthropic's Usage Policy.
- If Claude declines a request, we show you a clear message rather than forcing an answer.
- Rate limits keep the free beta fair and discourage automated misuse.
- Every plan is labelled as AI-generated, and we remind you to review it.
Your data
We don't store your idea or plan in a database. Anthropic processes API requests under its Commercial Terms and Privacy Policy; by default, API inputs and outputs aren't used to train Anthropic's models. See our privacy policy for the full picture.
ShipMVP is an independent product. We are not affiliated with, sponsored by or endorsed by Anthropic. “Claude” and “Anthropic” are trademarks of Anthropic, PBC.