Tools

Overview

Tools in FloTorch are MCP (Model Context Protocol) integrations that extend the capabilities of your AI agents. These tools allow agents to interact with external services, APIs, and data sources, enabling more sophisticated and context-aware responses.

Accessing the Tools Section

To access the Tools management interface:

1. Navigate to the Agents section in the left sidebar
2. Click on Tools (highlighted with a wrench and screwdriver icon)
3. You’ll be presented with the Tools management page

Tools Management Interface

The Tools page displays a comprehensive table of all configured tools in your workspace with the following information:

• Name: The unique identifier for each tool
• Type: Currently supports “MCP” (Model Context Protocol)
• URL: The endpoint URL where the tool is hosted
• Transport: The communication protocol (SSE or STREAM)
• Created At: Timestamp when the tool was created
• Actions: Options to edit or delete the tool

Creating a New Tool

Adding an MCP Tool

1. Click the ”+ Add MCP Tool” button in the top right corner
2. Fill in the required information:
   • Name*: A unique identifier for your tool (required)
   • Description: Optional description of the tool’s purpose
   • URL*: The endpoint URL where your MCP tool is hosted (required)
   • Transport: Select the communication protocol
     • SSE (Server-Sent Events): Real-time streaming communication
     • HTTP Stream: Standard HTTP streaming protocol
3. Headers: Add custom HTTP headers if required for authentication or configuration
4. Click Save to create the tool

Tool Configuration Best Practices

• Use descriptive names that clearly indicate the tool’s function
• Ensure the URL is accessible from your FloTorch workspace
• Choose the appropriate transport protocol based on your tool’s requirements
• Add necessary headers for authentication (API keys, tokens, etc.)

Managing Existing Tools

Editing a Tool

1. Locate the tool you want to edit in the tools table
2. Click the vertical ellipsis (⋮) icon in the Actions column
3. Select Edit from the dropdown menu
4. Modify the tool’s configuration as needed
5. Click Save to apply changes

Archiving and Unarchiving a Tool

Archiving hides a tool from the active list but keeps it available for dependencies. You can restore it later.

Archive a tool

1. In the left navigation, go to Tools.
2. In the tools table, open the Actions (three dots) menu for the tool and select Archive.
3. In the confirmation dialog, review the message: “Are you sure you want to archive this tool? You can restore it later if needed.”
4. Click Archive. The tool is archived and removed from the active list. Existing dependencies continue to work.
5. To view archived tools, open the top filters dropdown and select Archived.

Unarchive a tool

1. In the tools table, use the top filters dropdown and select Archived.
2. Open the Actions (three dots) menu for the tool and select Unarchive.

Deleting a Tool

Deleting a tool permanently removes the tool and all related dependencies. This action cannot be undone.

1. In the left navigation, go to Agents > Tools.
2. In the tools table, open the Actions (three dots) menu for the tool and select Delete.
3. Review the dependency list in the deletion modal.
4. Click Continue. A confirmation modal opens and asks you to type the tool name.
5. Enter the tool name and click Permanently Delete.

After confirmation, the tool and all listed dependencies are deleted permanently and cannot be recovered.

Search and Filtering

Use the search bar labeled “Search for a tool” to quickly find specific tools by name or other attributes. The search functionality helps you locate tools in workspaces with many configured integrations.

MCP Tool Integration

What is MCP?

Model Context Protocol (MCP) is a standardized way for AI models to interact with external tools and data sources. MCP tools in FloTorch enable your agents to:

• Access real-time data from external APIs
• Perform specific actions (file operations, database queries, etc.)
• Integrate with third-party services
• Extend agent capabilities beyond basic text generation

Supported Transport Protocols

• SSE (Server-Sent Events): Ideal for real-time, streaming data updates
• HTTP Stream: Standard HTTP streaming protocol for continuous data flow

Use Cases

MCP tools are particularly useful for:

• Data Integration: Connecting to databases, APIs, and external data sources
• File Operations: Reading, writing, and managing files
• Service Integration: Interacting with third-party services and APIs
• Real-time Updates: Receiving live data streams and notifications
• Custom Workflows: Implementing specialized business logic and processes

Security Considerations

• Ensure all tool URLs use HTTPS for secure communication
• Implement proper authentication through headers when required
• Regularly review and audit tool permissions and access
• Monitor tool usage and performance metrics

Troubleshooting

Common Issues

1. Tool Connection Failed: Verify the URL is accessible and the service is running
2. Authentication Errors: Check that required headers are properly configured
3. Transport Protocol Mismatch: Ensure the selected transport matches your tool’s implementation

Getting Help

If you encounter issues with tool configuration or integration:

1. Verify your tool’s MCP implementation follows the protocol specification
2. Check the tool’s logs for error messages
3. Ensure network connectivity between FloTorch and your tool endpoint
4. Contact support if issues persist

Using Tools in Agents

After configuring your MCP tools, you can integrate them with your AI agents to extend their capabilities:

Agent Configuration

1. Navigate to the Agent Builder section in the left sidebar
2. Select or create an agent you want to configure
3. In the agent configuration, look for the Tools section
4. Add your configured MCP tools to the agent’s toolset

Tool Integration Benefits

• Enhanced Capabilities: Agents can now perform actions beyond text generation
• Real-time Data: Access to live information from external sources
• Automated Workflows: Execute complex tasks through tool integrations
• Context Awareness: Tools provide additional context for better responses

Example Use Cases

• Data Analysis: Agents can fetch and analyze data from databases or APIs
• File Operations: Read, write, and manage files through file system tools
• Service Integration: Interact with third-party services and applications
• Custom Functions: Execute specialized business logic and processes