LangChain4j Tools

Since Camel 4.8

Both producer and consumer are supported

The LangChain4j Tools Component allows you to use function calling features from Large Language Models (LLMs) supported by LangChain4j.

Maven users will need to add the following dependency to their pom.xml for this component:

<dependency>
    <groupId>org.apache.camel</groupId>
    <artifactId>camel-langchain4j-tools</artifactId>
    <version>x.x.x</version>
    <!-- use the same version as your Camel core version -->
</dependency>

URI format

Producer
langchain4j-tools:toolSet[?options]
Consumer
langchain4j-tools:toolSet[?options]

Where toolSet can be any string to uniquely identify the endpoint

Configuring Options

Camel components are configured on two separate levels:

  • component level

  • endpoint level

Configuring Component Options

At the component level, you set general and shared configurations that are, then, inherited by the endpoints. It is the highest configuration level.

For example, a component may have security settings, credentials for authentication, urls for network connection and so forth.

Some components only have a few options, and others may have many. Because components typically have pre-configured defaults that are commonly used, then you may often only need to configure a few options on a component; or none at all.

You can configure components using:

  • the Component DSL.

  • in a configuration file (application.properties, *.yaml files, etc).

  • directly in the Java code.

Configuring Endpoint Options

You usually spend more time setting up endpoints because they have many options. These options help you customize what you want the endpoint to do. The options are also categorized into whether the endpoint is used as a consumer (from), as a producer (to), or both.

Configuring endpoints is most often done directly in the endpoint URI as path and query parameters. You can also use the Endpoint DSL and DataFormat DSL as a type safe way of configuring endpoints and data formats in Java.

A good practice when configuring options is to use Property Placeholders.

Property placeholders provide a few benefits:

  • They help prevent using hardcoded urls, port numbers, sensitive information, and other settings.

  • They allow externalizing the configuration from the code.

  • They help the code to become more flexible and reusable.

The following two sections list all the options, firstly for the component followed by the endpoint.

Component Options

The LangChain4j Tools component supports 5 options, which are listed below.

Name Description Default Type

configuration (common)

The configuration.

LangChain4jToolsConfiguration

bridgeErrorHandler (consumer)

Allows for bridging the consumer to the Camel routing Error Handler, which mean any exceptions (if possible) occurred while the Camel consumer is trying to pickup incoming messages, or the likes, will now be processed as a message and handled by the routing Error Handler. Important: This is only possible if the 3rd party component allows Camel to be alerted if an exception was thrown. Some components handle this internally only, and therefore bridgeErrorHandler is not possible. In other situations we may improve the Camel component to hook into the 3rd party component and make this possible for future releases. By default the consumer will use the org.apache.camel.spi.ExceptionHandler to deal with exceptions, that will be logged at WARN or ERROR level and ignored.

false

boolean

lazyStartProducer (producer)

Whether the producer should be started lazy (on the first message). By starting lazy you can use this to allow CamelContext and routes to startup in situations where a producer may otherwise fail during starting and cause the route to fail being started. By deferring this startup to be lazy then the startup failure can be handled during routing messages via Camel’s routing error handlers. Beware that when the first message is processed then creating and starting the producer may take a little time and prolong the total processing time of the processing.

false

boolean

autowiredEnabled (advanced)

Whether autowiring is enabled. This is used for automatic autowiring options (the option must be marked as autowired) by looking up in the registry to find if there is a single instance of matching type, which then gets configured on the component. This can be used for automatic configuring JDBC data sources, JMS connection factories, AWS Clients, etc.

true

boolean

chatModel (advanced)

Autowired Chat Model of type dev.langchain4j.model.chat.ChatModel.

ChatModel

Endpoint Options

The LangChain4j Tools endpoint is configured using URI syntax:

langchain4j-tools:toolId

With the following path and query parameters:

Path Parameters (1 parameters)

Name Description Default Type

toolId (common)

Required The tool id.

String

Query Parameters (11 parameters)

Name Description Default Type

tags (common)

Required The tags for the tools.

String

description (consumer)

Tool description.

String

exposed (consumer)

Whether the tool is automatically exposed to the LLM. When false, the tool is added to a searchable list and can be discovered via the tool-search-tool.

true

boolean

name (consumer)

Tool name.

String

parameters (consumer)

List of Tool parameters in the form of parameter.=. This is a multi-value option with prefix: parameter.

Map

bridgeErrorHandler (consumer (advanced))

Allows for bridging the consumer to the Camel routing Error Handler, which mean any exceptions (if possible) occurred while the Camel consumer is trying to pickup incoming messages, or the likes, will now be processed as a message and handled by the routing Error Handler. Important: This is only possible if the 3rd party component allows Camel to be alerted if an exception was thrown. Some components handle this internally only, and therefore bridgeErrorHandler is not possible. In other situations we may improve the Camel component to hook into the 3rd party component and make this possible for future releases. By default the consumer will use the org.apache.camel.spi.ExceptionHandler to deal with exceptions, that will be logged at WARN or ERROR level and ignored.

false

boolean

camelToolParameter (consumer (advanced))

Tool’s Camel Parameters, programmatically define Tool description and parameters.

CamelSimpleToolParameter

exceptionHandler (consumer (advanced))

To let the consumer use a custom ExceptionHandler. Notice if the option bridgeErrorHandler is enabled then this option is not in use. By default the consumer will deal with exceptions, that will be logged at WARN or ERROR level and ignored.

ExceptionHandler

exchangePattern (consumer (advanced))

Sets the exchange pattern when the consumer creates an exchange.

Enum values:

  • InOnly

  • InOut

ExchangePattern

lazyStartProducer (producer (advanced))

Whether the producer should be started lazy (on the first message). By starting lazy you can use this to allow CamelContext and routes to startup in situations where a producer may otherwise fail during starting and cause the route to fail being started. By deferring this startup to be lazy then the startup failure can be handled during routing messages via Camel’s routing error handlers. Beware that when the first message is processed then creating and starting the producer may take a little time and prolong the total processing time of the processing.

false

boolean

chatModel (advanced)

Autowired Chat Model of type dev.langchain4j.model.chat.ChatModel.

ChatModel

Spring Boot Auto-Configuration

When using langchain4j-tools with Spring Boot make sure to use the following Maven dependency to have support for auto configuration:

<dependency>
  <groupId>org.apache.camel.springboot</groupId>
  <artifactId>camel-langchain4j-tools-starter</artifactId>
  <version>x.x.x</version>
  <!-- use the same version as your Camel core version -->
</dependency>

The component supports 6 options, which are listed below.

Name Description Default Type

camel.component.langchain4j-tools.autowired-enabled

Whether autowiring is enabled. This is used for automatic autowiring options (the option must be marked as autowired) by looking up in the registry to find if there is a single instance of matching type, which then gets configured on the component. This can be used for automatic configuring JDBC data sources, JMS connection factories, AWS Clients, etc.

true

Boolean

camel.component.langchain4j-tools.bridge-error-handler

Allows for bridging the consumer to the Camel routing Error Handler, which mean any exceptions (if possible) occurred while the Camel consumer is trying to pickup incoming messages, or the likes, will now be processed as a message and handled by the routing Error Handler. Important: This is only possible if the 3rd party component allows Camel to be alerted if an exception was thrown. Some components handle this internally only, and therefore bridgeErrorHandler is not possible. In other situations we may improve the Camel component to hook into the 3rd party component and make this possible for future releases. By default the consumer will use the org.apache.camel.spi.ExceptionHandler to deal with exceptions, that will be logged at WARN or ERROR level and ignored.

false

Boolean

camel.component.langchain4j-tools.chat-model

Chat Model of type dev.langchain4j.model.chat.ChatModel. The option is a dev.langchain4j.model.chat.ChatModel type.

ChatModel

camel.component.langchain4j-tools.configuration

The configuration. The option is a org.apache.camel.component.langchain4j.tools.LangChain4jToolsConfiguration type.

LangChain4jToolsConfiguration

camel.component.langchain4j-tools.enabled

Whether to enable auto configuration of the langchain4j-tools component. This is enabled by default.

Boolean

camel.component.langchain4j-tools.lazy-start-producer

Whether the producer should be started lazy (on the first message). By starting lazy you can use this to allow CamelContext and routes to startup in situations where a producer may otherwise fail during starting and cause the route to fail being started. By deferring this startup to be lazy then the startup failure can be handled during routing messages via Camel’s routing error handlers. Beware that when the first message is processed then creating and starting the producer may take a little time and prolong the total processing time of the processing.

false

Boolean

Usage

This component helps to use function-calling features from LLMs so that models can decide what functions (routes, in case of Camel) can be called (i.e.; routed).

Consider, for instance, two consumer routes capable of query an user database by user ID or by social security number (SSN).

Queries user by ID
from("langchain4j-tools:userInfo?tags=users&description=Query database by user ID")
    .to("sql:SELECT name FROM users WHERE id = :#number");
Queries user by SSN
from("langchain4j-tools:userInfo?tags=users&description=Query database by user social security ID")
    .to("sql:SELECT name FROM users WHERE ssn = :#ssn");

Now, consider a producer route that receives unstructured data as input. Such a route could consume this data, pass it to a LLM with function-calling capabilities (such as llama3.1, Granite Code 20b function calling, etc) and have the model decide which route to call.

Such a route could receive questions in english such as:

  • "What is the name of the user with user ID 1?"

  • "What is the name of the user with SSN 34.400.96?"

Produce
from(source)
    .to("langchain4j-tools:userInfo?tags=users");

Tool Tags

Consumer routes must define tags that groups together. The aforementioned routes would be part have the users tag. The users tag has two routes: queryById and queryBySSN

Parameters

The Tool Input parameter can be defined as an Endpoint multiValue option in the form of parameter.<name>=<type>, or via the endpoint option camelToolParameter for a programmatic approach. The parameters can be found as headers in the consumer route, in particular, if you define parameter.userId=5, in the consumer route ${header.userId} can be used.

Producer and consumer example:
from("direct:test")
    .to("langchain4j-tools:test1?tags=users");

from("langchain4j-tools:test1?tags=users&description=Query user database by user ID&parameter.userId=integer")
    .to("sql:SELECT name FROM users WHERE id = :#userId");
Usage example:
List<ChatMessage> messages = new ArrayList<>();
        messages.add(new SystemMessage("""
                You provide information about specific user name querying the database given a number.
                """));
        messages.add(new UserMessage("""
                What is the name of the user 1?
                """));

        Exchange message = fluentTemplate.to("direct:test").withBody(messages).request(Exchange.class);

Using a specific Model

The Camel LangChain4j tools component provides an abstraction for interacting with various types of Large Language Models (LLMs) supported by LangChain4j.

Integrating with specific LLM

To integrate with a specific LLM, users should follow the steps described below, which explain how to integrate with OpenAI.

Add the dependency for LangChain4j OpenAI support:

Example
<dependency>
      <groupId>dev.langchain4j</groupId>
      <artifactId>langchain4j-open-ai</artifactId>
    <version>x.x.x</version>
</dependency>

Initialize the OpenAI Chat Language Model, and add it to the Camel Registry:

ChatLanguageModel model = OpenAiChatModel.builder()
                .apiKey("NO_API_KEY")
                .modelName("llama3.1:latest")
                .temperature(0.0)
                .timeout(ofSeconds(60000))
                .build();
context.getRegistry().bind("chatModel", model);

Use the model in the Camel LangChain4j Chat Producer

 from("direct:chat")
      .to("langchain4j-tools:test?tags=users&chatModel=#chatModel");

To switch to another Large Language Model and its corresponding dependency, replace the langchain4j-open-ai dependency with the appropriate dependency for the desired model. Update the initialization parameters accordingly in the code snippet provided above.

Handling no Tools Called

In some circumstances, the LLM may decide not to call a tool. This is a valid scenario that needs to be handled by application developers. To do so, developers can get the LangChain4jTools.NO_TOOLS_CALLED_HEADER from the exchange.

Tool Search Tool

The Tool Search Tool is a native feature that allows LLMs to discover and access tools dynamically without consuming the entire context window with tool definitions.

Overview

When you have many tools available, exposing all of them to the LLM in every request can:

  • Consume significant context window space

  • Reduce the space available for actual conversation

  • Potentially confuse the LLM with too many options

The Tool Search Tool solves this by allowing you to mark certain tools as "searchable" (non-exposed). These tools are not automatically sent to the LLM but can be discovered when needed.

Using the exposed Parameter

By default, all tools are exposed to the LLM (exposed=true). To make a tool searchable instead:

from("langchain4j-tools:queryBySSN?tags=users&description=Query user database by social security number&parameter.ssn=string&exposed=false")
    .to("sql:SELECT name FROM users WHERE ssn = :#ssn");

How It Works

  1. When you define tools with exposed=false, they are added to a searchable tool registry

  2. A native toolSearchTool is automatically exposed to the LLM when searchable tools exist

  3. The LLM can invoke toolSearchTool with tags to discover available tools

  4. The search results are returned to the LLM, which can then decide which tools to use

Example: Mixed Exposed and Searchable Tools

// This tool is immediately available to the LLM
from("langchain4j-tools:queryById?tags=users&description=Query user database by user ID&parameter.userId=integer")
    .to("sql:SELECT name FROM users WHERE id = :#userId");

// This tool is searchable but not immediately exposed
from("langchain4j-tools:queryBySSN?tags=users&description=Query user database by social security number&parameter.ssn=string&exposed=false")
    .to("sql:SELECT name FROM users WHERE ssn = :#ssn");

// Another searchable tool with different tags
from("langchain4j-tools:sendEmail?tags=users,email&description=Send email to a user&parameter.email=string&parameter.message=string&exposed=false")
    .to("smtp://mailserver");

In this example:

  • The queryById tool is immediately available to the LLM

  • The queryBySSN and sendEmail tools can be discovered by searching for tags like "users" or "email"

  • The LLM can use the toolSearchTool to find these additional capabilities when needed

Benefits

  • Reduced Context Usage: Only expose the most commonly used tools initially

  • Scalability: Support hundreds or thousands of tools without overwhelming the LLM

  • Dynamic Discovery: Let the LLM discover tools as needed based on the conversation

  • Better Organization: Group related tools by tags for easier discovery

Best Practices

When using the Tool Search Tool feature, consider the following best practices:

  • Tag Strategy: Use meaningful, hierarchical tags (e.g., "users", "users.admin", "database.users") to organize tools logically

  • Expose Common Tools: Keep frequently used tools exposed (exposed=true) and make specialized tools searchable (exposed=false)

  • Performance Considerations: While the search is efficient, having thousands of searchable tools may impact search performance. Consider grouping tools by functional area

  • LLM Guidance: In your system message, inform the LLM about the availability of the toolSearchTool and when to use it

  • Tag Naming: Use consistent, lowercase tag names without special characters for best compatibility

  • Tool Descriptions: Write clear, descriptive tool descriptions as they are returned in search results to help the LLM choose the right tool

  • Testing: Test your tool organization with real queries to ensure the LLM can discover and use tools effectively

Limitations

  • The LLM must be instructed to use the toolSearchTool - it won’t automatically know to search for tools

  • Search is based on exact tag matching - fuzzy matching or semantic search is not currently supported

  • The feature requires an LLM with good function-calling capabilities to work effectively

Back to top