agnost-ingestion

# Agnost Data Ingestion

Comprehensive guide for ingesting data into Agnost AI for analytics, monitoring, and insights. Covers the Conversation SDK for tracking AI interactions and the MCP SDK for Model Context Protocol server analytics.

> **Official docs:** [https://docs.agnost.ai](https://docs.agnost.ai)
> **API Endpoint:** `https://api.agnost.ai`
> **Dashboard:** [https://app.agnost.ai](https://app.agnost.ai)

## IMPORTANT: How to Apply This Skill

**Before implementing Agnost ingestion, follow this priority order:**

1. **Identify the use case**: Conversation tracking (AI chatbots, agents) or MCP server analytics
2. **Check SDK references** in the `references/` directory for detailed API
3. **Use provided code examples** as starting points
4. **Cite references** when explaining implementation details

---

## Quick Reference

### SDK Packages

| Use Case | Python | TypeScript/Node.js | Go |
|----------|--------|-------------------|-----|
| **Conversation/AI Tracking** | `pip install agnost` | `npm install agnostai` | N/A |
| **MCP Server Analytics** | `pip install agnost-mcp` | `npm install agnost` | `go get github.com/agnostai/agnost-go` |

### API Endpoints

| Endpoint | Method | Description |
|----------|--------|-------------|
| `/api/v1/capture-session` | POST | Create a new conversation/session |
| `/api/v1/capture-event` | POST | Record an event within a session |

---

## Conversation SDK (Recommended for AI Applications)

Use the Conversation SDK when building AI applications, chatbots, or agents that need to track user interactions, inputs, outputs, and performance metrics.

### Python Installation & Setup

```python
# Installation
pip install agnost
# or
uv add agnost

# Basic Setup
import agnost

# Initialize with your org ID (from dashboard)
agnost.init("your-org-id")
```

### TypeScript/Node.js Installation & Setup

```typescript
// Installation
npm install agnostai
// or
pnpm add agnostai

// Basic Setup
import * as agnost from "agnostai";

// In...