Chat Agent – Create

Getting started with creating your first chat agent on Flora AI

This article walks you through how to create your first chat agent on Flora AI by defining its “Identity” using the Base System Prompt. A well-designed identity helps your agent act consistently, match your brand’s voice, and respond effectively to users.

What is a Chat Agent’s Identity?

Your chat agent’s Identity is configured in the Base System Prompt on the AI Chat Agent details page.

This is where you tell the AI:

• Who it is (Role)

• What it can do (Skills)

• How it should sound (Personality)

• What it’s trying to achieve (Goal)

• How to think through problems (Chain of Thought)

• How to format answers (Output Format)

• What to avoid (Negative Rules)

• How good responses should look (Examples)

All of these together create an AI that matches your brand’s vibe and behaves predictably.

1. Define the Role

What it is:
The Role tells the agent who it is and how it should interpret user messages. This is the foundation of its identity.

How to write it:
Describe the agent’s position or function in one or two clear sentences.

Role example:

You are the customer service agent of an e-commerce website (www.xyz.com).

Use cases:

• “You are a sales assistant for a SaaS product.”

• “You are a technical support specialist for our mobile app.”

• “You are a virtual receptionist for a dental clinic.”

2. Specify the Skills

What it is:
Skills describe what the agent is capable of doing and which tools, data, or methods it should use.

How to write it:
List the main tasks the agent should perform and any resources it can rely on.

Skills example:

You need to help customers solve problems and answer their questions to the best of your ability.

You can expand this with more detail, such as:

• “Use the internal knowledge base to answer product questions.”

• “Guide users step-by-step through troubleshooting.”

• “Escalate complex issues by collecting details and summarizing them.”

3. Set the Personality

What it is:
Personality defines the tone, style, and communication approach of your agent.

How to write it:
Describe how the agent should sound in a few adjectives or a short sentence.

Personality example:

You are to provide professional and polite responses to users.

Other personality styles might be:

• “Friendly, casual, and upbeat.”

• “Formal, concise, and respectful.”

• “Empathetic, patient, and reassuring.”

Choose a style that matches your brand and audience.

4. Clarify the Goal

What it is:
The Goal tells the agent what its primary purpose is. This keeps responses focused and aligned with your business needs.

How to write it:
Use a clear, specific statement starting with “Goal: …”.

Goal example:

Goal: To assist customers with product-related inquiries and resolve any issues they may have.

You can tailor this to your use case, such as:

• “Goal: To help visitors find the right subscription plan and encourage sign-ups.”

• “Goal: To reduce support tickets by answering common questions accurately.”

5. Define the Chain of Thought

What it is:
Chain of Thought is a step-by-step process the agent should follow when handling user messages. It guides how the agent thinks and responds.

How to write it:
List the steps in order, from greeting to resolution.

Chain of Thought example:

1. Greet the customer and ask how you can assist them.

2. Listen to the customer's issue or question.

3. Access the internal knowledge base to find the relevant information.

4. Provide a clear and concise answer or solution to the customer.

5. Confirm with the customer if their issue has been resolved or if they need further assistance.

You can adapt these steps to your workflow, for example adding:

• “6. Offer related resources or suggestions if helpful.”

• “7. Ask if they would like to schedule an appointment.”

6. Choose the Output Format

What it is:
Output Format tells the agent how to structure and present its responses.

How to write it:
Specify the format (plain text, Markdown, JSON, etc.), length, and style.

Output Format example:

The response should be in plain text, easy to understand, and should not exceed three sentences. Use a friendly tone and avoid technical jargon.

Other examples:

• “Respond in Markdown with headings and bullet points where helpful.”

• “Return data as a JSON object with keys: answer, steps, and links.”

• “Use short paragraphs and numbered lists for instructions.”

7. Add Negative Rules

What it is:
Negative Rules define what the agent must not do. This helps prevent unwanted behavior and keeps responses safe and on-brand.

How to write it:
List clear “Do not…” or “Avoid…” rules.

Negative Rules example:

• Do not use negative or confrontational language, even if the customer is upset.

• Avoid providing overly technical explanations that may confuse the customer.

• Do not share personal opinions or off-topic comments.

• Keep responses concise; do not provide lengthy paragraphs that may overwhelm the customer.

• Refrain from making promises or guarantees that cannot be kept.

You can also add rules like:

• “Do not mention internal tools or systems by name.”

• “Do not provide medical, legal, or financial advice.”

8. Include Examples

What it is:
Examples show the agent what good responses look like. They are one of the most effective ways to shape behavior.

How to write it:
Provide a sample user question and the ideal agent response.

Example:

Customer Question:

I'm having trouble setting up my new device.

Chat Agent Response:

I'm here to help! Could you tell me where you're getting stuck in the setup process?

You can add multiple examples for common scenarios, such as:

• Pre-sales questions

• Troubleshooting steps

• Refund or cancellation requests

9. Full Base System Prompt Example

Below is a complete example combining all the elements into a single Base System Prompt you can adapt for your own agent.

Role:
Customer Support Specialist

Skills:
Problem-solving, product knowledge, empathy

Personality:
Professional, friendly, and patient

Goal:
To assist customers with product-related inquiries and resolve any issues they may have.

Chain of Thought:

1. Greet the customer and ask how you can assist them.

2. Listen to the customer's issue or question.

3. Access the internal knowledge base to find the relevant information.

4. Provide a clear and concise answer or solution to the customer.

5. Ask the customer if they would like to schedule an appointment.

Output Format:

The response should be in plain text, easy to understand, and should not exceed three sentences. Use a friendly tone and avoid technical jargon.

Negative Rules:

• Do not use negative or confrontational language, even if the customer is upset.

• Avoid providing overly technical explanations that may confuse the customer.

• Do not share personal opinions or off-topic comments.

• Keep responses concise; do not provide lengthy paragraphs that may overwhelm the customer.

• Refrain from making promises or guarantees that cannot be kept.

Example Interaction:

• Customer: I'm having trouble setting up my new device.

• Bot: I'm here to help! Could you tell me where you're getting stuck in the setup process?

10. Next Steps

1. Open your chat agent’s details page in Flora AI.

2. Locate the Base System Prompt section.

3. Fill in each part: Role, Skills, Personality, Goal, Chain of Thought, Output Format, Negative Rules, and Examples.

4. Save your changes and test the agent with a few sample questions to confirm it behaves as expected.

By carefully defining each of these elements, you create a chat agent that is consistent, helpful, and aligned with your brand.