# Introduction

Kaizen is an open-source AI automation agent that runs locally on your machine. Give it a task, and a team of specialized agents collaborates to get it done: researching, browsing, automating workflows, and learning from every run.

## Why Kaizen?

Most AI tools are stateless. They forget everything between sessions and repeat the same mistakes. Kaizen is different:

* **It automates, not just chats.** Kaizen browses the web, fills out forms, calls APIs, runs code, and connects to thousands of apps to complete real tasks end-to-end.
* **It learns.** After every run, insights and learnings are persisted to memory. The more you use it, the better it gets.
* **It's yours.** Runs entirely on your machine. No data leaves your environment except API calls to the model provider you choose.

## What Can Kaizen Do?

### Everyday Automation

| You say...                                                                        | Kaizen does...                                                                                                                           |
| --------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------- |
| "Take this photo of a recipe and add all the ingredients to my Instacart cart"    | Reads the image, extracts ingredients and quantities, opens Instacart in the browser, searches for each item, and adds them to your cart |
| "Every morning, check Hacker News for AI articles and send me a WhatsApp summary" | Creates a scheduled skill that runs daily, scrapes HN, filters for AI topics, summarizes the top stories, and messages you on WhatsApp   |
| "Find me flights to Tokyo under $800 for next month"                              | Searches multiple travel sites, compares prices, and presents the best options with booking links                                        |

### Business Automation

| You say...                                                                           | Kaizen does...                                                                                                            |
| ------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------- |
| "Search LinkedIn for senior Python developer jobs and classify them by salary range" | Browses LinkedIn, extracts job listings, classifies by salary band, and generates a structured report                     |
| "Every week, pull our sales data from the CRM and build a performance report"        | Creates a scheduled skill that connects to your CRM via Zapier, pulls the data, generates charts, and delivers the report |
| "Research our top 5 competitors and summarize what they shipped this month"          | Investigates each competitor's blog, changelog, and social media, then produces a competitive intelligence summary        |

### Creative & Technical

| You say...                                        | Kaizen does...                                                                                                                     |
| ------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------- |
| "Build me a weather dashboard for my city"        | Researches weather APIs, writes the code, tests it through a quality pipeline with visual review, and delivers a working dashboard |
| "Analyze this CSV and tell me what's interesting" | Reads the data, runs statistical analysis, generates visualizations, and presents findings with insights                           |

{% hint style="info" %}
**A note on AI outputs:** Kaizen is powered by large language models (LLMs), which are non-deterministic by nature. The same task may produce slightly different results each time. For important or sensitive workflows, we recommend reviewing outputs before acting on them.
{% endhint %}

## Key Features

* **Multi-agent pipeline** with specialized agents for routing, planning, executing, and reviewing
* **Skills system** for reusable automations that improve themselves over time
* **Scheduling** to run any skill on a cron schedule
* **Plugins** for custom code (Python, Node.js, TypeScript, Bash, Ruby)
* **Integrations** with WhatsApp, Zapier (8,000+ apps), Brave Search, and browser automation
* **Memory** that persists across sessions and improves with every run
* **Multi-model support** via OpenRouter: Claude, GPT, Gemini, DeepSeek, and more
* **Fully local** with secrets stored in an encrypted vault (AES-256-GCM)


---

# 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.kaizen-agent.ai/readme.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.