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

# Tools & Integrations

> Connect your agents to external APIs, MCP servers, and business systems

## Overview

eigi.ai agents become powerful when connected to your existing systems. Use Dynamic API Tools for REST integrations and MCP (Model Context Protocol) for advanced AI tool connections.

<Info>
  Already have your own OpenClaw, Harness, or custom OpenAI-compatible Responses
  API? Follow [Connect Your Own Harness](/features/connect-your-own-harness) for
  the simplest setup path.
</Info>

***

## Dynamic API Tools

### What Are API Tools?

API Tools let your agents interact with external services during conversations:

* Check order status in your database
* Book appointments in your calendar
* Look up customer information in your CRM
* Process payments through payment gateways
* Send emails or SMS notifications

### Creating an API Tool

<Steps>
  <Step title="Define the Endpoint">
    Specify the URL, HTTP method, and headers
  </Step>

  <Step title="Configure Parameters">Define what data the tool needs</Step>
  <Step title="Map Responses">Tell the agent how to use the response</Step>

  <Step title="Test the Tool">
    Validate it works before using in production
  </Step>

  <Step title="Assign to Agent">Add the tool to your agent's capabilities</Step>
</Steps>

***

## API Tool Configuration

### Endpoint Settings

| Setting         | Description                  | Example                             |
| --------------- | ---------------------------- | ----------------------------------- |
| **Name**        | Descriptive tool name        | "Check Order Status"                |
| **Description** | What the tool does (for LLM) | "Looks up order status by order ID" |
| **URL**         | API endpoint                 | `https://api.example.com/orders`    |
| **Method**      | HTTP method                  | GET, POST, PUT, DELETE              |
| **Headers**     | Request headers              | Authorization, Content-Type         |

### Authentication

Support for multiple auth methods:

<CardGroup cols={2}>
  <Card title="API Key" icon="key">
    Pass API key in header or query parameter
  </Card>

  <Card title="Bearer Token" icon="shield">
    JWT or OAuth bearer tokens
  </Card>

  <Card title="Basic Auth" icon="lock">
    Username and password authentication
  </Card>

  <Card title="Custom Headers" icon="code">
    Any custom authentication headers
  </Card>
</CardGroup>

***

## Parameter Types

### How Data Gets to Your API

<AccordionGroup>
  <Accordion title="Static Values" icon="lock">
    Fixed values that don't change:

    ```json theme={null}
    {
      "source": "voice_agent",
      "version": "1.0"
    }
    ```
  </Accordion>

  <Accordion title="LLM Parameters" icon="brain">
    Values extracted from conversation by AI:

    * "Get the order ID the customer mentioned"
    * "Extract the appointment date and time"
    * "Capture the email address provided"
  </Accordion>

  <Accordion title="User Input" icon="keyboard">
    Direct input from the caller:

    * DTMF digits entered
    * Specific prompted responses
  </Accordion>

  <Accordion title="Contact Variables" icon="user">
    Data from the phonebook contact:

    * `{{customer_id}}`
    * `{{account_number}}`
    * `{{email}}`
  </Accordion>
</AccordionGroup>

***

## Response Handling

### Processing API Responses

Configure how your agent uses the response:

```json theme={null}
// API Response
{
  "order_id": "ORD-12345",
  "status": "shipped",
  "tracking_number": "1Z999AA10123456784",
  "estimated_delivery": "2025-12-06"
}

// Agent says:
"Your order ORD-12345 has shipped! The tracking number is
1Z999AA10123456784 and it should arrive by December 6th."
```

### Error Handling

Define fallback behavior:

| Scenario         | Configuration                                                   |
| ---------------- | --------------------------------------------------------------- |
| **API Timeout**  | "I'm having trouble looking that up. Let me try again."         |
| **Not Found**    | "I couldn't find an order with that number. Can you verify it?" |
| **Server Error** | "Our system is temporarily unavailable. Can I take a message?"  |

***

## MCP (Model Context Protocol)

### What is MCP?

MCP is a standard for connecting AI models to external tools and data sources:

<CardGroup cols={2}>
  <Card title="Tool Discovery" icon="magnifying-glass">
    MCP servers automatically expose available tools to your agent
  </Card>

  <Card title="Rich Capabilities" icon="wand-sparkles">
    Access databases, file systems, specialized AI tools, and more
  </Card>

  <Card title="Standardized" icon="certificate">
    Works with any MCP-compatible server
  </Card>

  <Card title="Real-Time" icon="bolt">
    Persistent connections for fast tool execution
  </Card>
</CardGroup>

### MCP Server Types

| Type      | Description           | Use Case                      |
| --------- | --------------------- | ----------------------------- |
| **STDIO** | Standard input/output | Local tools, CLI integrations |
| **HTTP**  | REST-based servers    | Remote services, cloud tools  |

***

## Adding MCP Servers

### Configuration

<Steps>
  <Step title="Add Server">Provide server name and connection details</Step>

  <Step title="Configure Connection">
    Set up STDIO command or HTTP endpoint
  </Step>

  <Step title="Test Connection">Verify the server connects successfully</Step>
  <Step title="Discover Tools">See what tools the server provides</Step>
  <Step title="Assign to Agent">Enable the MCP server for your agent</Step>
</Steps>

### Connection Settings

```yaml theme={null}
# STDIO Server Example
name: "Database Tools"
type: stdio
command: "npx"
args: ["@your-org/db-mcp-server"]
env:
  DATABASE_URL: "postgresql://..."

# HTTP Server Example
name: "Calendar Integration"
type: http
url: "https://mcp.yourservice.com"
headers:
  Authorization: "Bearer your_token"
```

***

## MCP Performance

### Optimized Initialization

eigi.ai optimizes MCP for voice conversations:

<CardGroup cols={2}>
  <Card title="Connection Pooling" icon="diagram-project">
    Reuse connections across calls to reduce latency
  </Card>

  <Card title="Smart Caching" icon="database">
    Cache tool definitions and reduce startup time
  </Card>

  <Card title="Lazy Loading" icon="clock">
    Initialize servers only when needed
  </Card>

  <Card title="Health Monitoring" icon="heart-pulse">
    Automatic reconnection if servers disconnect
  </Card>
</CardGroup>

### Performance Metrics

| Metric              | Target                      |
| ------------------- | --------------------------- |
| **Connection Time** | \< 500ms for cached servers |
| **Tool Execution**  | \< 200ms overhead           |
| **Reconnection**    | Automatic within 5 seconds  |

***

## Tool Usage in Conversations

### How Agents Use Tools

The LLM decides when to use tools based on conversation context:

**Customer:** "What's the status of my order 12345?"

**Agent (internally):**

1. Recognizes need for order lookup
2. Calls `check_order_status` tool with order\_id="12345"
3. Receives response with status and tracking
4. Formulates natural response

**Agent (to customer):** "I found your order! It shipped yesterday and should arrive by Friday. Would you like the tracking number?"

### Silent Tool Calls

<Info>
  Tools are called silently in the background. Customers only hear the agent's
  natural response—never technical details about tool execution.
</Info>

***

## Multiple Tools

### Combining Capabilities

Agents can use multiple tools in a single conversation:

```
1. Check customer identity (CRM lookup)
2. Verify account status (Billing API)
3. Check order history (Orders API)
4. Book follow-up appointment (Calendar tool)
5. Send confirmation email (Email tool)
```

### Tool Chaining

Tools can be called sequentially based on conversation flow:

```
Customer asks about upgrading →
  → Check current plan (API) →
  → Get available upgrades (API) →
  → Apply upgrade if confirmed (API) →
  → Send confirmation (Email)
```

***

## Testing Tools

### Validation

Before going live, test your tools:

<CardGroup cols={2}>
  <Card title="Direct Testing" icon="flask">
    Test API calls with sample data
  </Card>

  <Card title="Mock Responses" icon="clone">
    Simulate responses for edge cases
  </Card>

  <Card title="Conversation Testing" icon="comments">
    Full conversation tests with real tool calls
  </Card>

  <Card title="Error Simulation" icon="bug">
    Test error handling and fallbacks
  </Card>
</CardGroup>

***

## Monitoring

### Tool Performance

Track how your tools perform:

| Metric           | Description                    |
| ---------------- | ------------------------------ |
| **Call Volume**  | How often each tool is used    |
| **Success Rate** | Percentage of successful calls |
| **Latency**      | Average response time          |
| **Errors**       | Error types and frequencies    |

### MCP Server Status

Monitor MCP server health:

* Connection status (connected/disconnected)
* Last activity timestamp
* Error logs
* Tool availability

***

## Best Practices

<Tip>
  **Clear Descriptions**: Write tool descriptions that help the LLM understand
  when to use them.
</Tip>

<Tip>
  **Handle Errors Gracefully**: Always configure fallback responses for API
  failures.
</Tip>

<Tip>
  **Test Edge Cases**: Verify behavior with missing data, timeouts, and errors.
</Tip>

<Warning>
  Never expose sensitive credentials in tool configurations. Use environment
  variables or secure credential storage.
</Warning>

***

## Troubleshooting

<AccordionGroup>
  <Accordion title="Tool not being called">
    * Check the tool description—LLM may not understand when to use it - Verify
      the tool is assigned to the agent - Test with explicit prompts mentioning
      the tool's function
  </Accordion>

  <Accordion title="MCP server not connecting">
    * Verify connection settings (URL, command, arguments) - Check server logs for
      errors - Ensure required environment variables are set - Test the server
      independently first
  </Accordion>

  <Accordion title="Slow tool responses">
    * Optimize your API endpoints - Consider caching frequently accessed data -
      Use connection pooling for MCP servers - Monitor and address timeout issues
  </Accordion>
</AccordionGroup>