> For the complete documentation index, see [llms.txt](https://vibemx.gitbook.io/neuronai-studio/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://vibemx.gitbook.io/neuronai-studio/getting-started/quickstart-first-workflow.md).

# Quickstart: First Workflow

Run a pre-built workflow template, inspect the execution trace, and understand the studio runtime.

## What you will build

A single-agent chat workflow using the `basic-agent-chat` template — the simplest end-to-end workflow in the studio.

## Prerequisites

* [Installation](/neuronai-studio/getting-started/installation.md) completed
* At least one agent exists (or use a template that creates one automatically)

## Step 1 — Browse templates

Navigate to **Templates**:

```
/neuronai-studio/templates
```

Filter by **Workflows** and find **Basic Agent Chat**. Click **Use Template**.

The installer creates any required agents, remaps graph references, and opens the workflow editor.

## Step 2 — Inspect the graph

The canvas shows a minimal flow:

```mermaid
flowchart LR
    Start[Start] --> Agent[Agent] --> Stop[Stop]
```

Click the **Agent** node to see its configuration: which agent runs, the message template (`{{input}}`), and the output state key.

## Step 3 — Run the test harness

Open the **Test** panel (workflow chat harness). Type a message and send it.

The runtime:

1. Sets `input` in workflow state from your message
2. Executes nodes in graph order
3. Streams step events and tokens via SSE
4. Persists a trace record

## Step 4 — Inspect the trace

After the run completes, open **Traces** for this workflow. Click the latest trace to see:

* Per-step timeline (start → agent → stop)
* Input and output payloads for each step
* Total duration and any errors

## Optional: Autonomous loop with attachments

For the full autonomous agent pattern (loop + tools + multimodal input):

1. Open **Templates** and install **Autonomous Lead Qualification**
2. Open the workflow editor and inspect the **Loop** → **Agent** → **Condition** subgraph
3. In the **Test** panel, attach a PDF or image and send an initial message
4. Watch SSE events: `loop_iteration`, `tool_call`, `tool_result`
5. If the agent needs more data, reply when the workflow pauses at the **Human** node

The agent retains thread memory across iterations via `__studio_thread_id`. See [Autonomous agents in workflows](/neuronai-studio/workflows/overview.md#autonomous-agents-in-workflows).

## Optional: RAG upstream of an agent

1. Create a **Knowledge Base** under `/neuronai-studio/knowledge-bases` and ingest a document
2. Install the **Support RAG HITL** template (or add a RAG node manually before an Agent node)
3. Confirm the Agent message references `{{ rag_context.context }}`
4. Run the test harness — trace steps include `rag_query` metadata

See [AI Nodes — RAG](/neuronai-studio/workflows/ai-nodes.md#rag).

## Next steps

* [Workflow Overview](/neuronai-studio/workflows/overview.md) — concepts and node types
* [Canvas Editor](/neuronai-studio/workflows/canvas-editor.md) — build custom graphs
* [State & Conditions](/neuronai-studio/workflows/state-and-conditions.md) — branch on workflow data
* [Templates](/neuronai-studio/workflows/templates.md) — try `lead-qualification` or `support-rag-hitl`


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://vibemx.gitbook.io/neuronai-studio/getting-started/quickstart-first-workflow.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.