Anthropic Claude Function Calling and Tools Integration

Tools4AI is an open-source project that showcases a 100% Java implementation of a Large Action Model (LAM). It functions as a Language Model (LLM) agent and is designed to integrate AI with enterprise Java applications. The project facilitates converting natural language prompts into actionable behaviors, which can be particularly useful in automating workflows and enhancing interaction with AI systems.

Anthropic Integration

Get the key from here -
https://docs.anthropic.com/claude/reference/getting-started-with-the-api

Inside tools4ai.properties you need to set the key , if you prefer you can also pass the key with system parameter -DclaudeKey=

#set it here or use -DclaudeKey parameter
claudeKey=

Calling actions

AnthropicActionProcessor processor = new AnthropicActionProcessor();
        String postABook = "post a book harry poster with id 189 the publish date is 2024-03-22 and the description is about harry who likes poster its around 500 pages  ";
        String result = (String)processor.processSingleAction(postABook);
        Assertions.assertNotNull(result);
        String success = TestHelperOpenAI.getInstance().sendMessage("Look at this message - "+result+" - was it a success? - Reply in true or false only");
        log.debug(success);
        Assertions.assertTrue("True".equalsIgnoreCase(success));

Full code is here

Spring Integration

SpringAnthropicProcessor processor = new SpringAnthropicProcessor(context);
(String)processor.processSingleAction(postABook);
        Assertions.assertNotNull(result);

Features

Tools4AI offers several advantages for developers looking to integrate AI into their Java applications. Here are some of the key benefits:

Language Compatibility: Being built entirely in Java, Tools4AI seamlessly integrates with existing Java applications, allowing developers to leverage AI capabilities without needing to introduce additional languages or technologies into their stack.

Function Calling and Actionable Prompts: It supports function calling directly from natural language prompts, which means it can execute Java methods, REST API calls, or even shell scripts based on user inputs. This reduces the complexity and code necessary to interpret user commands and execute actions.

Multi AI Voting Mechanism: Tools4AI can consult multiple AI models to determine the best course of action for a given input. This feature enhances decision-making accuracy by aggregating insights from various sources.

Flexible Integration: It works with popular AI platforms like OpenAI and local AI models, offering flexibility depending on the application's needs and privacy requirements.

Autonomous Agent Capabilities: The project allows developers to build autonomous agents that can perform a range of tasks autonomously, from fetching data to processing complex queries, enhancing automation within software applications.

Ease of Use: The ability to map natural language prompts to specific actions without manual intervention makes Tools4AI a user-friendly option for developers looking to enhance their applications with AI without extensive additional coding.

These features make Tools4AI a robust tool for any developer aiming to integrate AI functionalities into enterprise-level Java applications efficiently.