# Quick Start

### 1. Create your first agent

1. Go to **Agents → Create Agent**.
2. Fill in the basic details:
   * **Name**: A clear, user‑friendly name (for example, “Sales Assistant” or “Support Bot”).​
   * **Agent Image** and **Image Description** (optional): Visual and short persona for your agent.​
   * **Welcome Message**: The first message users see when they open a chat.​
3. Choose how the agent thinks:
   * **Provider**: Select the AI model provider from the dropdown.
   * **Temperature**: Lower for precise, factual answers; higher for more creative responses.
   * **Always On RAG Search**: Turn on if you want the agent to always look up answers in its Knowledge Base.
4. Click **Create** to save the agent.​

Your new agent now appears in the **Agents** list, where you can search for it and open **Manage Agent** to configure it further.​

***

### 2. Create a Knowledge Base

1. Go to **Knowledge Bases → Create / Edit Knowledge Base**.​
2. Enter:
   * **Title**\*: A descriptive name for the Knowledge Base (for example, “Product Docs” or “Instagram Library”).​
   * **Description**\*: A short explanation of what content it contains and how it should be used.​
3. Click **Save**.​

Now you have an empty Knowledge Base ready for content.​

***

### 3. Add knowledge from documents and Instagram

1. In your Knowledge Base, click **Add Knowledge Resources**.​
2. Choose the type of content you want to import:
   * **Document**
     * Upload or connect files such as manuals, FAQs, guides, or reports.​
     * These documents become searchable so agents can answer using your exact content.​
   * **Instagram**
     * Connect your Instagram content so the system can use posts, captions, and **transcripts from your reels** where available.​
     * This helps agents understand your brand tone and social campaigns and generate answers in the same style.​

Any agent linked to this Knowledge Base can now use these resources in its responses.

***

### 4. Configure how your agent behaves (prompts)

1. Open **Agents → Manage Agent → Prompt Configuration** for your agent.
2. Choose a **Prompt Type** to edit:
   * **Base Prompt**: Main behavior and rules the agent follows in every conversation.​
   * **Welcome Message Prompt**: Controls the first message when a new chat starts.​
   * **Always On RAG Search Prompt**: Instructions for how to use the Knowledge Base when RAG is always on.​
   * **Welcome Back Message Prompt**: Message shown to returning users.​
   * **User Memory Tool Prompt**: How the agent reads and uses saved user memory.​
3. Use **Prompt Versions** to track changes:
   * Increment **Minor versions** for small refinements (wording, clarifications).​
   * Use a new **Major version** for bigger behavior or logic changes.​
   * Update the **Change Log** each time so your team knows what changed and why.​
4. Use **Optimise Prompts** and **Prompt Evaluation** to test and improve prompts based on sample questions and real conversations.​

***

### 5. Test your agent

1. Go to **Try Chat** to talk to your agent in a safe preview.
2. Ask real user questions and confirm:
   * Answers follow your rules and tone.
   * Outputs match the required format (for example, JSON schema or specific sections).​
3. Open **Conversations** to review past chats, spot misunderstandings, and decide where prompts or Knowledge Bases need updates.​

Make adjustments in **Agent Info**, **Prompt Configuration**, and **Knowledge Bases**, then retest in **Try Chat**.

***

### 6. Publish and share your agent

1. Open the **Publish** tab for your agent.​
2. Configure how users will see and use it:
   * **Try Asking Questions Count**: Number of example questions shown in the chat UI.​
   * **Chat Widget**: Copy the embed code (with your **bot-id** and styling options) into your website or app.​
   * **Standalone Chat Page**: Use the “Open Page” link to share a ready‑to‑use chat URL with your team or customers.​
3. Add tracking and integrations:
   * **Google Analytics Measurement ID** to send usage events to your GA property.​
   * **Integrations** like DoubleTick and Sendpulse to use the agent in channels such as WhatsApp or campaigns.​
4. Click **Save** to apply your publish settings and make the latest configuration live.​

Once these steps are complete, you have a live Brainstormer agent powered by your own Knowledge Bases and finely tuned prompts, ready to support your users around the clock.

<br>


---

# Agent Instructions: 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:

```
GET https://brainstormer.gitbook.io/brainstormer-docs/quick-start.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.