> ## Documentation Index
> Fetch the complete documentation index at: https://docs.p2hs.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Build with P2HS AI

> Use prompts and codegen runs to create or modify a P2HS project.

P2HS AI turns a project request into a tracked codegen run. The run can estimate cost, collect required user inputs, execute generation, validate the result, and produce a revision for review.

## Write an effective request

Describe the outcome, the important user roles, the data the application needs, and any integrations. For example:

```text theme={null}
Build a customer portal with sign-in, customer profiles,
subscription status, a support request form, and an admin view.
Use the project’s admitted authentication, forms, and record capabilities.
```

Avoid asking for credentials or embedding provider secrets in prompts or generated client code.

## Run lifecycle

1. Estimate the run when an estimate is available.
2. Create the run with the project context and requested change.
3. Supply required user inputs or capability approvals.
4. Execute and monitor status and timeline events.
5. Review changes and the resulting revision.
6. Preview and deploy after validation.

See [Codegen lifecycle](/content/ai/codegen-lifecycle) for the public abstraction and check the dashboard for the current credit treatment.

For build failures, waiting runs, and missing changes, see [Build troubleshooting](/content/support/troubleshooting/builds).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.