# Quickstart

#### How It Works

MCP (Model Context Protocol) lets AI tools like Claude, ChatGPT, and Gemini access your business systems directly. Once connected, you can ask questions and give commands in plain English.

**Read and Write to Kick**

Your AI can query transactions, generate reports, and update bookkeeping — all with your approval.

| What You Ask                                                                                                                                      | What Happens                                                                                                |
| ------------------------------------------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------- |
| <p><strong>Query and analyze</strong><br><code>Show me all transactions over $5,000 from last quarter and break them down by category.</code></p> | Your AI pulls the data from Kick and gives you an instant analysis.                                         |
| <p><strong>Create automation</strong><br><code>Create a rule to categorize all Amazon transactions as Office Supplies.</code></p>                 | Your AI drafts the rule, shows you what it will do, and waits for your approval before creating it in Kick. |

**Connect Multiple Tools in One Conversation**

Add your CRM, project management software, or property tools alongside Kick. Your AI can pull from all of them at once.

| What You Ask                                                                                                                                                                                                                                     | What Happens                                                                                          |
| ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------- |
| <p><strong>Cross-system analysis (Kick + CRM)</strong><br><code>Compare revenue by client in Kick to deal size in HubSpot. Show me which clients are most profitable relative to contract value.</code></p>                                      | Your AI queries both systems, combines the data, and delivers one answer.                             |
| <p><strong>Multi-platform workflow (Kick + Project Management)</strong><br><code>For all projects closed in Linear last quarter, pull actual expenses from Kick and compare to budgeted amounts. Which projects went over budget?</code></p>     | Your AI identifies completed projects from Linear, pulls expenses from Kick, and highlights overruns. |
| <p><strong>Property management + accounting (Kick + Property Software)</strong><br><code>Pull rent payments from Kick for January and compare to expected rent from lease agreements in Buildium. Show me which units are delinquent.</code></p> | Your AI matches payments to leases across both systems and flags delinquencies.                       |

Want more examples? Browse real prompts and video walkthroughs at Kick [MCP Use Case Library](/ai/kick-mcp/mcp-use-case-library.md)

***

#### Where to Start

**Step 1: Connect Kick to Your AI Tool**

Pick the AI you already use - Claude, ChatGPT, Gemini, Cursor, Perplexity, Microsoft Copilot Studio, or any MCP-compatible tool.

Each connector guide has step-by-step screenshots. Setup takes 5 minutes.

→ [View all Kick connectors](/ai/kick-mcp/connectors.md)

**Step 2: Add Your Other Tools (Optional)**

If your CRM, project tracker, or property management software supports MCP, add it to the same AI tool. Now your AI can query multiple systems in one conversation.

***

#### First prompts with MCP

Once connected, you can ask your AI tool to:

* Find and analyze transactions: "Show me all transactions over $5,000 from last quarter"
* Generate financial reports: "Create a profit and loss report for January 2026"
* Review client data: "What are the top 5 expenses for \[client name] this year?"
* Work with accounts: "List all asset accounts and their current balances"
* Create accounting rules: "Create a rule to categorize all Amazon transactions as Office Supplies"
* Analyze trends: "Compare Q1 and Q2 revenue across all entities"

Most tools are read-only by default. If the AI tool needs to make changes (like creating a rule or updating a transaction), it will ask for your explicit confirmation first.

***

#### The Benefits

* **No more context switching:** One conversation spans all your tools. No more tabbing between Kick, your CRM, your project tracker, and three spreadsheets.
* **Live data, not stale exports:** Your AI pulls current data every time. No exports, no copy-paste, no outdated CSVs.
* **Plain English, not code:** You don't need to be a developer. Ask questions in natural language. The AI figures out which systems to query and how to combine the results.
* **Quick setup:** Connecting Kick to your AI tool takes 5 minutes. No technical knowledge required - just follow the screenshots in the connector guide.
* **You stay in control:** When the AI needs to update data (create a rule, categorize a transaction), it shows you what it plans to do and waits for your approval.

***

#### Need More Help?

* **Getting connected**: See connector-specific guides for step-by-step troubleshooting
* **Ideas for what to ask**: Browse the use case library for prompts and videos
* **Questions about MCP**: Contact <support@kick.co>


---

# 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://docs.kick.co/ai/kick-mcp/mcp-quickstart.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.