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

> ## Agent Instructions
> For API operations, fetch https://gateway.firsttouch.ai/api/public/openapi.json for the complete input/output schemas, internal references, and x-mcp-annotations. Endpoint Markdown intentionally omits generated schema fragments. Select /api/public/tools/<tool_name> using the exact underscore-separated name. Send requests to https://gateway.firsttouch.ai; the canonical document currently omits servers. Start at https://docs.firsttouch.com/api-reference/agent-start.md.

# Personalize LinkedIn Connection Notes

> Import an audience, personalize a connection request with variables or AI, preview notes, and review before sending.

Use one Flow Plan to prepare a personalized LinkedIn connection note for each prospect. You can control the wording with variables or enable AI messaging to generate a note from available context.

## Before You Start

* A Team Admin can create the CSV Audience and configure the Flow Plan.
* The assigned sender needs a connected, ready LinkedIn account. Sending a note requires LinkedIn Premium or Sales Navigator.
* A connection note can contain up to **300 characters**. Check the rendered note, including inserted values.
* Prospects need a usable personal LinkedIn identity before the connection request can execute.

See [Social Integration](https://docs.firsttouch.com/WebApp/Social/social-Integration.md) for account setup and [LinkedIn Actions](https://docs.firsttouch.com/WebApp/Social/Social-Orchestration.md) for action requirements.

## 1. Import Your Audience

Open **Audiences**, choose **New audience > CSV file**, and upload your file. Map the supported columns, review validation, and choose **Create flow**. You can also save the audience first and create a flow from it later.

For this workflow, include prospect names and known personal LinkedIn profile URLs. Use the [CSV example and mapping guide](https://docs.firsttouch.com/WebApp/Audiences.md#csv-notes) for the full requirements.

This workflow personalizes a template or generates notes from supported data. The standard CSV importer does not import arbitrary custom fields or a prewritten note per row; see [CSV fields and personalization](https://docs.firsttouch.com/WebApp/Audiences.md#csv-fields-and-personalization).

## 2. Prepare the Data Your Note Needs

For a simple first-name and company-name note, use the available saved fields. If the note needs enriched company details, review the Flow Plan's enrichment settings. If it needs research, place a [Research Agent](https://docs.firsttouch.com/WebApp/Flows/Agents/researcher.md) before the connection request on the same path.

For a value you already store in HubSpot, select the available HubSpot property. See [Variables and Personalization](https://docs.firsttouch.com/WebApp/Flows/variables-and-personalization.md) for exact syntax and data-source differences.

## 3. Add a Connection Request

In Flow Builder, add a connection-request step, review the assigned LinkedIn sender, and enter a note. Choose the messaging mode that fits your goal.

### Authored Note with Variables

Turn **AI messaging** off and write the final wording. Insert variables from the editor's variable menu:

```text theme={null}
Hi {ft.flow.prospect_first_name}, I'd like to connect and learn more about your work at {ft.flow.company_name}.
```

The wording stays the same while each prospect's available values are inserted.

### AI-Generated Note

Turn **AI messaging** on before using instruction blocks. For example:

```text theme={null}
Hi {ft.flow.prospect_first_name}, [Write one short reason to connect based on the available company context. Do not invent facts. Keep the complete note under 300 characters.]
```

Add relevant variables or an earlier research output through the selector to supply context. Generated wording still needs preview and review. With AI messaging off, text in square brackets is ordinary text and can be sent literally.

## 4. Preview Different Prospects

Open **Preview**, choose a contact, and inspect the rendered note. Repeat with another contact, then check a prospect with missing optional information or a long name.

Confirm that:

* Names and company details are correct.
* Research-based statements are supported by the available findings.
* Missing data does not leave an incomplete sentence or an unsupported claim.
* The complete note is within 300 characters.

If a value is unavailable, change the wording or data setup before publishing. See [Missing Values and Preview](https://docs.firsttouch.com/WebApp/Flows/variables-and-personalization.md#missing-values-and-preview).

## 5. Review Approval and Timing

Require human approval if a person should review each request before it sends. Review the action delay, sender Work Schedule, and daily capacity. Check both the **Wait for acceptance** window and the separate withdrawal timer.

Approval tasks are managed in [Tasks](https://docs.firsttouch.com/WebApp/Tasks.md). Approval does not bypass account readiness, scheduling, or capacity requirements.

## 6. Publish and Enroll

Review the source, sender assignment, exclusions, and enrollment mode, then publish when ready. Use manual enrollment if you want to select the first prospects yourself; an automatic mode can enroll eligible contacts according to its settings.

Follow [Publishing and Routing](https://docs.firsttouch.com/WebApp/Flows/publishing.md) and [Flow Enrollments](https://docs.firsttouch.com/WebApp/Flows/enrollments.md), then monitor requests in the [Outreach Queue](https://docs.firsttouch.com/WebApp/Outreach-Queue.md).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.