### Install MindCache

Source: https://context7_llms

Install the MindCache library using npm. This is the first step to integrating persistent memory into your AI applications.

```bash
npm install mindcache
```

--------------------------------

### Basic MindCache React App Setup

Source: https://context7_llms

A minimal example of setting up a full AI chat application using MindCache's React components. This includes the `MindCacheProvider` and `MindCacheChat` components.

```tsx
import { MindCacheProvider, MindCacheChat } from 'mindcache';

// Full AI chat app in ~15 lines
function App() {
  return (
    <MindCacheProvider
      ai={{ provider: 'openai', model: 'gpt-4o', keyStorage: 'localStorage' }}
    >
      <MindCacheChat />
    </MindCacheProvider>
  );
}
```

--------------------------------

### Generate LLM Tools and System Prompt

Source: https://context7_llms

Illustrates how to automatically generate tools for AI SDKs like Vercel AI SDK using `get_aisdk_tools`. It also shows how to retrieve the system prompt.

```typescript
// Auto-generate Vercel AI SDK tools
const tools = mc.get_aisdk_tools();

// For regular keys: write_<key>
// For document keys: write_<key>, append_<key>, insert_<key>, edit_<key>

const prompt = mc.get_system_prompt();
```

--------------------------------

### Implement Undo/Redo Functionality

Source: https://context7_llms

Demonstrates how to use undo and redo operations, both on a per-key basis and globally. Accessing the global history is also shown.

```typescript
mc.undo('key');     // Per-key undo
mc.undoAll();       // Global undo
mc.getGlobalHistory(); // Get history
```

--------------------------------

### Initialize MindCache in Cloud (Real-Time Sync) Mode

Source: https://context7_llms

Initialize MindCache for real-time cloud synchronization. This requires specifying cloud instance details and a token endpoint. Await synchronization after initialization.

```typescript
import { MindCache } from 'mindcache/cloud';
const mc = new MindCache({ cloud: { instanceId: 'xxx', tokenEndpoint: '/api/ws-token' } });
await mc.waitForSync();
```

--------------------------------

### Store and Retrieve Text Values

Source: https://context7_llms

Demonstrates storing and retrieving simple text values using `set_value` and `get_value`. This is the default behavior for storing data.

```typescript
mc.set_value('name', 'Alice');
mc.get_value('name'); // 'Alice'
```

--------------------------------

### Store and Manipulate Document Values

Source: https://context7_llms

Shows how to store and update collaborative documents. Documents can be created with `set_value` or `set_document`, and updated using character-level edits.

```typescript
// Create with set_value
mc.set_value('notes', '# My Notes', { type: 'document' });

// Or use convenience method
mc.set_document('notes', '# My Notes');

// Update (uses diff automatically)
mc.set_value('notes', 'Updated content');

// Character-level edits
mc.insert_text('notes', 0, 'New text');
```

--------------------------------

### Initialize MindCache in Offline (IndexedDB) Mode

Source: https://context7_llms

Initialize MindCache using IndexedDB for local persistence. This mode allows data to persist across browser sessions. Ensure to await synchronization.

```typescript
const mc = new MindCache({ indexedDB: { dbName: 'my-app' } });
await mc.waitForSync();
```

--------------------------------

### MindCache React Hooks Usage

Source: https://context7_llms

Shows how to use various MindCache React hooks to access the MindCache instance, build custom chat UIs, and manage synchronization with GitHub.

```tsx
import { useMindCacheContext, useClientChat, useLocalFirstSync } from 'mindcache';

// Access MindCache instance
const { mindcache, isLoaded, hasApiKey, setApiKey, getModel } = useMindCacheContext();

// Build custom chat UI
const { messages, sendMessage, isLoading, streamingContent } = useClientChat();

// Sync to GitHub
const { status, save, load } = useLocalFirstSync({ mindcache, gitstore: { owner, repo, token } });
```

--------------------------------

### Initialize MindCache in Memory Mode

Source: https://context7_llms

Initialize MindCache in its default 'Memory Only' mode. This mode stores data in RAM and is suitable for simple use cases or testing.

```typescript
import { MindCache } from 'mindcache';
const mc = new MindCache();
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.