Toolkits & Indexes

​

Introduction to Toolkits

Toolkits in ELITEA are modular integrations that connect your agents to external platforms and internal services. Each toolkit provides an interface for configuring connection parameters (such as API keys, URLs, or credentials) and setting options specific to the target service. Once configured, toolkits can be assigned to agents, empowering them to automate tasks across different systems and domains.

---

​

Navigate the Toolkits & Indexes Menu

The Toolkits & Indexes menu is accessible from the main platform navigation. Upon entering this section, you will see a dashboard listing all available toolkits along with their configuration status. Main elements:

• Toolkit cards: Each card provides a brief description, current status, and toolkit type.
• Search and filter: Locate a toolkit using the search bar or filter by type.
• Create Toolkit button: Use the + Create button to add a new integration or configure an existing one.
• View toggle: Switch between Card List View and Table View using the toggle in the top-right corner.
• Pinning: Pin frequently used toolkits to keep them at the top of the list.

---

​

Create a New Toolkit

Add new integrations to the platform by creating a new toolkit through the Toolkits menu.

​

Start Toolkit Creation

• From the Toolkits & Indexes page: Select the + Toolkit button at the top of the page.
• From any page: Open the global Create dropdown in the navigation bar and select Toolkit.

​

Select a Toolkit Type

Choose the desired toolkit type from the categorized list. Toolkit types are organized by category to help you find the integration you need. Use the search bar to locate a specific toolkit type by name.

​

Configure Toolkit Details

Once you have selected a toolkit type, configure the following fields: Main configuration fields:

​

Save Your Toolkit

Select Save (top right) to create the toolkit. It will appear in your toolkits dashboard, where you can access its detailed configuration page to modify settings or adjust tool selection as needed.

---

​

Toolkit Categories & Descriptions

The platform supports a wide range of toolkits, organized by category. This table lists the available categories and toolkit types.

---

​

Manage Toolkit Configuration

The toolkit detail page provides a comprehensive interface for configuring, testing, and monitoring your toolkits. Access it by selecting any toolkit from the Toolkits dashboard.

​

Configure & Test a Toolkit

The toolkit detail page is the primary interface for configuring toolkit settings, managing tool selection, and testing toolkit functionality in real time. The page header includes a Back button, the toolkit name, and a controls dropdown in the top-right area. The layout is split into two main areas:

• Left panel — the configuration form, including toolkit settings, tool selection, a Run History button above the form, and an Indexes accordion section when indexing is supported for the toolkit.
• Right panel — the testing area, which shows an empty state before a tool is selected, Test Settings after a tool is selected, and Run Results after a tool run.

On the toolkit detail page, you can:

• Edit toolkit details: Edit the name, description, credentials, and other toolkit-specific settings in the left panel.
• Select tools: In the Tools section, turn on only the tools your agent needs.
• Update advanced settings: Update PgVector configuration or embedding model as your indexing needs change.
• Save or discard: Save becomes active once you make changes. Changing the embedding model while indexes exist triggers a warning first, since it can make those indexes non-operational.
• View run history: Select the Run History button above the form to view past test runs.
• Use the controls menu: Use the three-dot menu (⋮) for Copy Link, Pin/Unpin, and Delete. Export and Fork are listed but currently unavailable.
• View authors: Shown in the header when viewing a public toolkit.
• Edit Raw JSON: Supported toolkits let you switch the form to Raw JSON editing via the view toggle in the left panel.

​

Make Tools Available by MCP

When configuring toolkit tools, you may see a checkbox labeled Make Tools Available by MCP in the Tools section. This option controls whether the selected tools can be accessed through the Model Context Protocol (MCP) interface.

• Purpose: Turns on toolkit tools to be exposed and accessible via MCP servers.
• Location: Appears in the Tools configuration section when selecting toolkit tools.
• When to turn on:
  • You plan to use this toolkit’s tools through an MCP server.
  • You want to make these tools available to external MCP clients.
  • You are building MCP-based integrations that need access to these tools.

​

Test a Toolkit

The right panel contains the toolkit testing workflow:

• Before a tool is selected, the panel shows an empty state.
• After a tool is selected, the panel shows Test Settings.
• After you run a tool, the panel switches to Run Results, where you can review the output and use the back button to return to the settings view.

For comprehensive testing instructions and best practices, see How to Test Toolkit Tools. Steps to test a toolkit:

1. Select a tool: Use the Tool dropdown to choose which toolkit action you want to test. Type in the search box to filter available tools and select from the options.
2. Select a model and adjust settings (optional): Choose the LLM model from the model dropdown (for example, gpt-4o). To fine-tune the response, select the Model Settings icon (⚙️) next to the model selector. The available settings vary by model:
   • Reasoning Models (e.g. GPT-5.1)
   • Standard Models (e.g. GPT-4o)

   Setting	Option	Description
   Reasoning	Low	Fast, surface-level reasoning with concise answers and minimal steps
   	Medium (default)	Balanced reasoning with clear explanations and moderate multi-step thinking
   	High	Deep, thorough reasoning with detailed step-by-step analysis (may be slower)
   Max Completion Tokens	Auto (default)	System automatically sets the token limit to 4096 tokens
   	Custom	Manually set a specific token limit; interface shows remaining tokens available. An error is shown if the value exceeds the model’s maximum output tokens

   Setting	Option	Description
   Creativity	1	Highly focused and deterministic outputs
   	2	Mostly focused with slight variation
   	3 (default)	Balanced between focus and creativity
   	4	More varied and creative responses
   	5	Maximum creativity and diversity
   Max Completion Tokens	Auto (default)	System automatically sets the token limit to 4096 tokens
   	Custom	Manually set a specific token limit; interface shows remaining tokens available. An error is shown if the value exceeds the model’s maximum output tokens

3. Provide required parameters (if prompted): Enter any inputs required by the selected tool.
4. Run the tool: Select the Run Test button and review the results in the output area. The Run Test button appears after selecting a tool.

---

​

Shared Toolkit Credentials

Shared toolkits in team projects can show warnings when each user must provide a matching private credential, or when a credential change may affect other team members.

---

​

Indexes Section

Toolkits that include indexing tools expose an Indexes accordion in the left configuration panel. Indexing ingests documents into a vector store so agents can search and retrieve content using semantic similarity. When indexes exist, the accordion lists index cards. Each card shows the collection name, indexed/total document count, skipped document count, current status, and last run date. From each card you can:

• Open — select the card to open the index detail page.
• Reindex — rebuild the index; a confirmation is required before existing data is replaced.
• Delete — remove the index (requires confirmation).
• Open in new tab — open the index detail page in a separate browser tab.

To create a new index, select the + Index button at the bottom of the list. For detailed instructions, see How to Create and Use Indexes.

---

​

View Toolkit Run History

The Run History page provides a complete audit trail of all past test executions performed in the toolkit’s Test Settings panel. How to access Run History:

1. Open any saved toolkit.
2. Go to the toolkit’s detail page and select the View run history button. This opens the dedicated Run History page for that toolkit.
3. Use the back navigation (breadcrumb or back button) to return to the toolkit’s configuration tab.

Run History layout: The Run History page uses the shared RunHistoryContainer layout:

• Left panel — sortable list of all past test runs. Select any column header to sort:

  Column	Description
  Date	Timestamp of the run in dd-MM-yyyy, hh:mm a format
  Version	Toolkit version used for the run
  Duration	Total execution time of the run

  Select a row to load the conversation replay in the right panel. Each row also has a three-dot menu (⋮) with Share link and Delete.
• Right panel — full conversation replay of the selected test run.

---

​

Use Toolkits in Your Workflows

Once a toolkit is configured, use it across different areas of the platform to enhance your workflows. In Agents

1. Go to the Agents menu.
2. Select or create an agent.
3. In the agent’s configuration, add the desired toolkit from the list of available toolkits.
4. Configure any agent-specific toolkit options, if prompted.
5. Select Save.

In Pipelines

1. Go to the Pipelines menu.
2. Select or create a pipeline.
3. In the pipeline configuration, add the desired toolkit to turn on automated workflow steps.
4. Configure pipeline-specific toolkit parameters as needed.
5. Select Save.

In Chat

1. Open a Chat session.
2. Access the toolkit selector within the chat interface.
3. Select one or more toolkits to make their capabilities available during the conversation.
4. Use natural language to interact with the toolkit’s features.

---

​

Best Practices

Secure Your Credentials

Use dedicated credentials for each integration, store them only in the platform credential manager, and avoid sharing private credentials across users or workspaces.

Test Before Use

Test each turned-on tool in the Test Settings panel before assigning the toolkit to agents, pipelines, or chat workflows.

Keep Toolkits Updated

Review toolkit settings whenever credentials rotate, service URLs change, available tools change, or indexing configuration is updated.

Minimal Permissions

Turn on only the tools and external-service permissions required for the intended workflow.

---

​

Troubleshooting

Failed Connection

Verify the selected credential, confirm any required URLs or project identifiers, and check that the target service is reachable from your environment.

Toolkit Not Appearing

Refresh the Toolkits page, clear active filters or search terms, and verify that you have permission to view or create that toolkit type.

Agent Cannot Use Toolkit

Confirm the toolkit is saved, properly configured, and explicitly attached to the agent, pipeline, or chat session where you expect to use it.

Indexes Section Not Available

If the Indexes section does not appear or cannot be used in your toolkit:

1. Verify indexing tools are turned on — Confirm the Index data tool is selected in the toolkit’s Tools section.
2. Check AI configuration — Confirm that PgVector and an Embedding Model are configured in AI Configuration.
3. Save and refresh — Save the toolkit after making changes, then refresh the page so the Indexes accordion can load with the updated configuration.

For further assistance, contact your platform administrator.

​

Support Contact

If you encounter issues not covered in this guide or need additional assistance with toolkit management, see Contact Support for detailed information on how to reach the ELITEA Support Team.

---