> ## Documentation Index
> Fetch the complete documentation index at: https://docs.iblueprint.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Get started with iBlueprint: your first Blueprint
> Sign up for iBlueprint, connect an AI provider, build your first Blueprint with a prompt node, run it, and review the execution log.
This guide walks you through the fastest path to a working Blueprint in iBlueprint. You will sign up for an account, connect an AI provider, create a Blueprint with a prompt node, execute it, and inspect the result in the execution log.
Go to [app.iblueprint.ai](https://app.iblueprint.ai) and click **Sign up**. Enter your email address and a password of at least six characters, then follow the confirmation link sent to your inbox.
If you prefer passwordless sign-in, use **Sign in with magic link** on the login page — iBlueprint will email you a one-time link instead.
Before you can run a prompt node, you need at least one AI provider key.
1. Open **Settings** from the sidebar, then select **AI Connections**.
2. Click **Add provider key**.
3. Choose a provider — **OpenAI**, **Anthropic**, **Google**, or **Ollama** — and paste your API key.
4. Give the key a name (for example, `OpenAI production`) and toggle **Set as default** if you want this provider selected by default in new nodes.
5. Click **Save**.
iBlueprint stores your key encrypted and scoped to your account. Other users cannot access it.
1. Click **New Blueprint** in the left sidebar.
2. Give your Blueprint a name — for example, `Summarize article`.
3. Click **Create**. The visual editor opens with an empty canvas.
1. Click **Add node** (or the **+** button on the canvas) and select **Prompt**.
2. In the node settings panel that opens on the right, choose your AI provider and model.
3. Write your prompt in the **System prompt** or **User message** field. For a first test, try something simple:
```
Summarize the following text in two sentences: {{input}}
```
The `{{input}}` placeholder is a [variable](/blueprints/variables) that you can supply at runtime.
4. Click **Save node**.
You can chain multiple nodes together by clicking the output handle at the bottom of any node and dragging to the input handle of the next. Output from earlier nodes flows automatically into later ones.
Click **Run** in the top-right toolbar. If your Blueprint has variables, iBlueprint shows an input panel — fill in the values and click **Execute**.
You can also run any Blueprint from the CLI. First install the CLI:
```bash theme={null}
npm install -g @iblueprint/cli
```
Then execute your Blueprint by ID:
```bash Execute a Blueprint theme={null}
iblueprint execute
```
```bash Execute with input variables theme={null}
iblueprint execute --inputs '{"input": "Paste your article text here."}'
```
```bash List all Blueprints theme={null}
iblueprint list
```
```bash Get Blueprint details theme={null}
iblueprint get
```
When execution starts, the CLI prints the execution ID:
```
Execution started, ID: exec_abc123
```
Use that ID to check the status at any time:
```bash theme={null}
iblueprint status exec_abc123
```
Back in the web app, open the **Executions** tab on your Blueprint. Each row shows the execution timestamp, status, and duration. Click any row to expand the full trace — you can inspect the input and output for every node in the chain, along with token usage and latency.
Executions are stored and searchable. You can replay any previous execution with the same inputs by clicking **Re-run** on its detail page.
## Next steps
Now that your first Blueprint is running, explore what else you can build:
Add API, function, MCP, and conditional nodes to your chain.
Pass dynamic data into any node at runtime.
Full reference for all CLI commands: list, get, execute, status.
Connect additional providers and manage per-key rate limits.