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

# Quickstart

> Run your first Deepline SDK workflow safely: inspect, pilot, review, then scale. Use it for a first run, result review, and safe scaling.

Job: take one real GTM record, run a workflow, and see whether Deepline returns useful enrichment before you scale to a list.

Start with a prebuilt workflow. Write code only when you need custom routing, custom waterfalls, or scheduled automation.

## Before you start

You need the Deepline CLI installed from the main [Quick Start](/docs/quickstart), plus an approved workspace.

## 1. Authenticate

```bash theme={null}
deepline auth register
deepline auth status --json
```

You should see `connected: true` and the workspace you are using.

## 2. Inspect the workflow

Before running anything, ask Deepline what the workflow expects:

```bash theme={null}
deepline plays describe prebuilt/person-linkedin-to-email --json
```

Look for required inputs and example output. If you are using Claude Code, ask it to summarize the schema in plain English.

## 3. Pilot one real record

Use a real lead you are allowed to enrich. Placeholder people may return no result.

```bash theme={null}
deepline plays run prebuilt/person-linkedin-to-email \
  --input '{"linkedin_url":"https://www.linkedin.com/in/real-person/"}' \
  --watch
```

Good outcome: the run finishes and returns verified contact data or a clear no-result reason.

## 4. Scale to a list

Once one record looks right, move to a CSV-capable play and pilot a few rows:

```bash theme={null}
head -n 4 leads.csv > /tmp/leads-pilot.csv

deepline plays run docs-examples/sdk-v2/lead-email-waterfall.play.ts \
  --csv /tmp/leads-pilot.csv \
  --watch

deepline runs export <run-id> --out /tmp/leads-pilot-enriched.csv
```

Inspect the output before running the full file.

## 5. Create a custom workflow

Use a custom workflow when your process has business logic:

* search once a week and send new matches to Salesforce
* run a different enrichment path for VPs vs directors
* try providers in your preferred waterfall order
* score accounts and export a chart

```ts theme={null}
import { definePlay } from 'deepline';

type SearchInput = {
  query: string;
  owner_email: string;
};

export default definePlay(
  'weekly-new-accounts',
  async (ctx, input: SearchInput) => {
    const search = await ctx.tools.execute({
      id: 'company_search',
      tool: 'test_rate_limit',
      input: { key: input.query },
      description: 'Find matching companies.',
    });

    const routing = {
      owner_email: input.owner_email,
      matched: search.status === 'SUCCESS',
    };

    return { routing };
  },
);
```

Check custom workflow files before running them:

```bash theme={null}
deepline plays check weekly-new-accounts.play.ts
```

This validates the workflow without starting a run.

## Ask Claude Code

```txt theme={null}
Read the Deepline SDK quickstart.
Inspect prebuilt/person-linkedin-to-email.
Pilot one real record.
If the output looks right, help me run a 3-row CSV pilot.
Do not run the full file until I review the pilot output.
```

<Tip>
  Use `plays check` whenever Claude Code writes or edits a play. It catches SDK
  syntax and play-shape problems before money or time gets spent.
</Tip>