Skip to main content

​
What is a Blueprint?

A Blueprint is a reusable AI workflow made from connected nodes. Each node performs a step, passes its output forward, and can be configured with inputs, models, tools, or human review. A Blueprint packages the workflow so it can be tested, reused, forked, and shared.

​
Key concepts

  • Nodes are the individual workflow steps. A node has a type, configuration, inputs, and an output.
  • Chains connect node outputs to downstream inputs so work moves through the workflow in a predictable order.
  • Variables let you supply different runtime values without editing the workflow. Use clear names and document the expected format.
  • Versions preserve changes over time so you can test and review updates before using them broadly.

​
Browse public Blueprints

  1. Open Public Libraries and select Blueprints, or go directly to iblueprint.ai/blueprints.
  2. Search with Search blueprints….
  3. Use Categories… and the sort control. The current sort options are Last Updated, Date Created, and Name.
  4. Review a published card’s description, visibility, categories, creator, and rating.

​
Fork a Blueprint

Select Fork blueprint on a public card to copy the workflow into your workspace. Forking gives you a separate version to inspect and adapt while leaving the public Blueprint unchanged. Sign in when Blueprint requests it.

​
Blueprint lifecycle

  1. Draft: create or fork a workflow, configure its nodes, and test its behavior.
  2. Published: share a reviewed version with the intended audience. Record what changed and what inputs, permissions, and dependencies it expects.
Continue versioning the workflow as it evolves. Treat publication as a release step, not as a substitute for testing.

​
Build reliable Blueprints

  • Give every node a clear purpose and define the expected input and output.
  • Make model, tool, knowledge, and human-review dependencies explicit.
  • Use safe test data and verify failure, timeout, and handoff behavior.
  • Review permissions and data handling before sharing or publishing.
  • Fork public workflows before making changes so the source remains intact.
Public Blueprints are starting points. Test a fork with representative, non-sensitive data before connecting it to production systems or consequential decisions.
Current Public Libraries view Current Blueprint Public Libraries view showing the Blueprints entry and search. The Public Libraries page is the starting point for finding published Blueprints and other reusable assets.