> For the complete documentation index, see [llms.txt](https://docs.gallabox.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.gallabox.com/chat-ai-agent/migrating-from-v1-agents-to-v2.md).

# Migrating from v1 agents to v2

### Step 1: Open the AI Agents & Bots listing page

Navigate to the unified agents/bots list to begin creating the new agent.

<figure><img src="/files/gg1iDg81d1eHAGfU4ex1" alt=""><figcaption></figcaption></figure>

***

### Step 2: Start creating a new Chat AI Agent

Click the agent-creation control and choose the agent type.

<figure><img src="/files/MkWGSAdrpuF64gwdQmMj" alt=""><figcaption></figcaption></figure>

***

### Step 3: Choose industry, use case, and name the agent

This step spans three wizard screens.

**3a — Industry selection:**

<figure><img src="/files/7ZKi9QneggGaQAj6LzP9" alt=""><figcaption></figcaption></figure>

**3b — Use case selection (choose "Build from scratch"):**

<figure><img src="/files/DAT4Cb1GaC3yGXRkQXCW" alt=""><figcaption></figcaption></figure>

**3c — Agent name and medium:**

<figure><img src="/files/Ufn27RCUtCX8UapJAmdV" alt=""><figcaption></figcaption></figure>

> **Note:** The wizard auto-fills a suggested agent name (e.g. "Education - Build from scratch") based on your industry + use-case choices; rename it as needed. Medium defaults to WhatsApp.

***

### Step 4 & 5: Open the v1 bot and locate Role, Goal, and Instructions

Open the v1 bot directly at its flow URL and review its Home tab, where Role, Goal, and Instructions are all visible together.

<figure><img src="/files/jBd5ERvHTaqcz8inwO5V" alt=""><figcaption></figcaption></figure>

> **Note:** In v1, Role/Goal/Instructions live together on the **Home** tab rather than separate tabs.

***

### Step 6: Paste Role/Goal/Instructions into the new agent, save, and test

Paste the copied content into the new v2 agent's Instructions tab and verify behavior using the built-in Live Testing panel.

<figure><img src="/files/RLc2rFlrxwMmeJFAaoXQ" alt=""><figcaption></figcaption></figure>

> **Note:** The v2 agent has a live "Live testing" chat panel built directly into the Instructions tab, useful for immediately validating the migrated Role/Goal/Instructions (this also covers the final Step 12 verification).

***

### Step 7: Review the v1 bot's custom actions

Open the v1 bot's Actions tab to see which actions need to be recreated.

<figure><img src="/files/Bi6c5RoR7Ln1PUxu5eD5" alt=""><figcaption></figcaption></figure>

> **Note:** v1's "My Actions" includes `assign_to_bot`, `assign_to_team`, `assign_to_user`, `booking_details`, and `get_details`, alongside System Actions `assign_to` and `get_current_date_time`.

***

### Step 8: Recreate the custom actions in the new v2 agent

<figure><img src="/files/l0JfmQeO0sT6bEkeHJ1I" alt=""><figcaption></figcaption></figure>

***

### Step 9: Identify files to migrate from the v1 bot

<figure><img src="/files/NwP56ZgJdy1kXB27lJ0z" alt=""><figcaption></figcaption></figure>

***

### Step 10: Create/link a Knowledge Base in the new agent

<figure><img src="/files/BPIWK5PfEeuUtuycc7pC" alt=""><figcaption></figcaption></figure>

> **Note:** v2 uses a dedicated **Knowledge Base** object (up to 3 can be attached per agent) rather than a flat file list. The "Dubai Housekeeping Knowledge Base" was created and linked here.

***

### Step 11: Configure Settings (AI Model)

<figure><img src="/files/QVACdWb1Wm7DbGzy7Nfz" alt=""><figcaption></figcaption></figure>

> **Note:** Under "AI Model & Document Intelligence": AI Provider is locked to **v2**, AI Model is set to **gpt-4o**, and LLM Temperature is **0.10**.