> For the complete documentation index, see llms.txt.
Skip to main content

Check out Port for yourself ➜ 

Craft a prompt

Choose a use-case​

After connecting your Git provider, you will be prompted to choose an initial use-case that Port will build for you.

Port offers several pre-built options for common use-cases, each with an example prompt you can use as-is or edit to add more context.

For example, selecting "Engineering Intelligence" generates a prompt like:

I want to measure AI coding-assistant adoption across my organization,
with visibility at the user, team, and org levels.

Provide context​

After submitting your prompt, Port AI assistant plans the implementation and asks clarifying questions to better understand your environment, such as:

  • Which AI coding assistant(s) are you tracking adoption for?
  • Do you want to grade teams against adoption targets?
  • How are your teams organized in Port or your identity provider?

These questions help Port create the right components for your specific setup. The assistant will not suggest connecting additional tools at this stage - that happens later, based on suggestions in the chat.

What Port creates for you​

Once you confirm the plan, Port sets up the initial structure of your platform:

  • Blueprints - data models representing your services, environments, and other resources. Learn more about the default blueprints.
  • Dashboard pages - pre-built visualizations for your selected use-case.
  • Workflows - automations your team can trigger for common tasks.

The AI assistant will detail exactly what was built and suggest your next steps.

It's always available via the button and can be used to change or add components at any time. It supports different modes of interaction - you can ask questions about your catalog, plan changes before making them, or build new components directly.

To learn more about the AI assistant, refer to the documentation.

Next step​

Once the AI assistant finishes the initial setup, you land on the services page in your platform, with the AI assistant chat still open on the right side. The chat walks you through whatever is left to get real data flowing - connecting additional tools, linking users to their profiles, verifying that key entities exist, and any other setup step specific to your use-case.

You can return to this conversation at any time using the button and browsing your chat history.

With your platform set up, proceed to set up your context lake.