> ## Documentation Index
> Fetch the complete documentation index at: https://docs.trackyard.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Using Trackyard with AI Agents

> Conversational music search and download via Claude, Cursor, and other MCP-compatible agents

Once the Trackyard MCP server is installed, you can search and download music conversationally through your AI agent. No code required.

***

## Basic Workflow

<Steps>
  <Step title="Describe what you need">
    **You:** "I need upbeat music for a 30-second Instagram Reel about travel"
  </Step>

  <Step title="The AI searches Trackyard">
    **Claude:** "I'll search Trackyard for upbeat travel music..."

    \[Presents 3-5 top results with titles, artists, BPM, moods]
  </Step>

  <Step title="Choose a track">
    **You:** "Download the second one as a 30-second clip"
  </Step>

  <Step title="The AI downloads and saves">
    **Claude:** "Downloaded 'Summer Wanderlust' as a 30-second clip → `travel_music.mp3`"
  </Step>
</Steps>

***

## Example Conversations

### Simple Search and Download

```
You: "Find me chill lo-fi music for a coding stream"

Claude: "I'll search Trackyard for chill lo-fi music..."

[Results:]
1. Late Night Code - Lo-Fi Cafe (120 BPM, Chill, Relaxed)
2. Focus Flow - Beat Lab (95 BPM, Ambient, Minimal)
3. Terminal Sessions - Code Beats (110 BPM, Lo-Fi, Calm)

You: "Download the first one"

Claude: "Downloaded 'Late Night Code' (2:42) → late_night_code.mp3"
```

***

### Smart Trimming

```
You: "Find upbeat music for a product reveal video"

Claude: [presents 5 results]

You: "Download #3 as a 22-second clip with the big hit at 12 seconds"

Claude: "Downloaded 'Digital Sunrise' as a 22-second clip with the musical peak aligned to 12 seconds → product_reveal_music.mp3"
```

***

### Batch Operations

```
You: "Find 3 different tracks for travel videos and download 15-second clips of each"

Claude: [searches, presents 3 tracks]

"Downloaded 15-second clips:
1. Summer Wanderlust → track1.mp3
2. Open Road → track2.mp3
3. Horizon Chaser → track3.mp3"
```

***

## Advanced Use Cases

<AccordionGroup>
  <Accordion title="Filter by specific criteria" icon="filter">
    **You:** "Find electronic music with no vocals, 120-140 BPM, high energy"

    **Claude:** \[searches with filters: `has_vocals: false`, `min_bpm: 120`, `max_bpm: 140`, `energy_level: "high"`]
  </Accordion>

  <Accordion title="Compare multiple tracks" icon="scale-balanced">
    **You:** "Show me 5 options for a corporate training video, then I'll pick one"

    **Claude:** \[presents 5 tracks with metadata, lets you audition via preview URLs, then downloads your choice]
  </Accordion>

  <Accordion title="Match music to specific scenes" icon="film">
    **You:** "I need tense music for a chase scene, around 90 seconds, building in intensity"

    **Claude:** \[searches with moods: "Tense", "Suspenseful", "Building" and downloads a 90-second clip]
  </Accordion>

  <Accordion title="A/B test variants" icon="flask">
    **You:** "Give me 3 different vibes for a product demo: upbeat, minimal, and cinematic"

    **Claude:** \[searches 3 times with different queries, downloads 3 clips]
  </Accordion>
</AccordionGroup>

***

## Tips for Best Results

<Tip>
  **Be specific:** "upbeat electronic for tech demo" > "upbeat music"
</Tip>

<Tip>
  **Mention the use case:** Context helps the AI infer genre, mood, and energy
</Tip>

<Tip>
  **Ask for previews:** "Show me previews first" lets you audition before downloading
</Tip>

<Tip>
  **Use natural language:** The AI understands "the big hit should land at 12 seconds" without technical jargon
</Tip>

***

## What the MCP Server Can Do

The Trackyard MCP server exposes these capabilities to AI agents:

<CardGroup cols={2}>
  <Card title="Search" icon="magnifying-glass">
    Natural language queries with optional filters (genre, mood, BPM, vocals, energy)
  </Card>

  <Card title="Download Full Tracks" icon="download">
    Get the complete audio file with perpetual license
  </Card>

  <Card title="Smart Trimming" icon="scissors">
    Auto-select best segments for specific durations
  </Card>

  <Card title="Hit Point Alignment" icon="bullseye">
    Choreograph musical peaks to specific offsets
  </Card>

  <Card title="Preview Streaming" icon="headphones">
    Listen to 30-second previews before downloading
  </Card>

  <Card title="Account Info" icon="user">
    Check credit balance and rate limits
  </Card>
</CardGroup>

***

## Limitations

<Warning>
  **Credit consumption:** Each search and download consumes credits, just like the REST API. The AI will warn you when credits are low.
</Warning>

<Warning>
  **Rate limits apply:** The MCP server is subject to the same rate limits as direct API calls (10 req/min on free tier).
</Warning>

***

## Next Steps

<CardGroup cols={2}>
  <Card title="Setup Guide" icon="gear" href="/mcp-server/setup-guide">
    Install the MCP server if you haven't yet
  </Card>

  <Card title="What is MCP?" icon="circle-info" href="/mcp-server/what-is-mcp">
    Learn more about the Model Context Protocol
  </Card>

  <Card title="API Reference" icon="book" href="/api-reference/search">
    See what's happening under the hood
  </Card>

  <Card title="Use Cases" icon="lightbulb" href="/use-cases/social-media-content-farms">
    Get inspired by real-world workflows
  </Card>
</CardGroup>