Tool usage
Tool use is an AI system's invocation of external software functions, services, or other capabilities to obtain information or perform actions.
Build an AI Agent (From Scratch)
A tool call is a model-generated specification requesting invocation of a particular external tool with particular arguments.
- How it is used
- Enables an LLM agent to request external information, computation, or actions during an iterative task.
- How deep it goes
- Works through schemas, Python dispatch, call IDs, result messages, and schema-design guidelines; stops short of requiring an agent loop for a single call.
- Examples
- Contrasts a direct capital-city answer with a calculator call for 1234 × 5678, then shows Python execution and a returned tool message.
- What you can do after
- Build it from scratch by defining a callable schema, dispatching generated calls, and feeding linked execution results back to the model.
Read how Build an AI Agent (From Scratch) explains tool usage →
AI Agents and Applications
A tool call is a structured model response that requests invocation of a named tool with supplied arguments.
- How it is used
- Enables an agent to select external capabilities and sequence calls as a user’s information needs unfold.
- How deep it goes
- Works through registration, message routing, execution, and debugging, then identifies a built-in ReAct alternative to manual graph construction.
- Examples
- Traces Cornwall beach searches with multiple queries and extends the travel assistant to check weather forecasts.
- What you can do after
- Build it from scratch with a tool-execution loop, then apply LangChain binding and LangGraph routing to a multi-tool agent.
Introduction to Generative AI, Second Edition
Tool use is an AI system capability in which the system invokes external software, services, or functions to perform actions or obtain information.
- How it is used
- Enables AI agents to obtain information or take actions through external tools while answering requests and completing multistep tasks.
- How deep it goes
- Explains structured calls, returned results, training approaches, and repeated calls at a high level without implementing an API.
- Examples
- Uses web search for current news, sample calculator and translation calls, and a spreadsheet-to-review-rating task.
- What you can do after
- Evaluate/understand when an agent needs tools or chained calls and why external access requires safeguards.
Read how Introduction to Generative AI, Second Edition explains tool usage →
AI for Everyday IT
Function calling is a language-model capability that enables a model to request invocation of defined software functions or external services by producing structured arguments.
- How it is used
- Connects natural-language requests to application logic and external services through predictable, machine-processable function arguments.
- How deep it goes
- Explains definitions, JSON flow, execution boundaries, database workflow, and security controls, but stops short of building a complete application.
- Examples
- Uses SQL generation and checking, currency conversion, image extraction, meeting-note processing, and HR data cleanup as practical examples.
- What you can do after
- Build it from scratch by defining functions, parsing model JSON, validating requests, executing approved actions, and returning results.
Reinforcement Learning from Human Feedback
Tool use is a model capability in which generated interactions invoke external tools or services to perform tasks.
- How it is used
- Extends language models with external information, precise computation, and real-world actions through structured requests.
- How deep it goes
- Explains orchestration, training objectives, formatting, masking, and multistep reasoning, while deferring detailed constrained-decoding mechanics.
- Examples
- Uses live-information search, file organization, pi calculation, movie search, Fibonacci code execution, and multistep web search.
- What you can do after
- Build it from scratch by implementing a model-call, external-execution, tool-result, continuation loop.
Read how Reinforcement Learning from Human Feedback explains tool usage →
Build AI-Enhanced Web Apps
Tool calling is a technique in which an AI model emits a structured request for an external tool or function, receives the execution result, and incorporates it into its response.
- How it is used
- Extends AI applications with access to external resources, specialized data, and actions beyond a model’s training scope.
- How deep it goes
- Traces the structured-request and execution cycle through the application server and SDK, including a less-flexible prompt-injection alternative.
- Examples
- Demonstrates weather retrieval and tool-call messages, with booking systems, e-commerce, and automation agents as use cases.
- What you can do after
- Apply tool calling with the Vercel AI SDK by defining custom tools, registering them, and handling their responses.
Build Python Web Apps with Streamlit
Tool use is the capability of an AI system to invoke external software functions or services while solving a task.
- How it is used
- Enables agents to complete personalized tasks by interacting with APIs, databases, and real-world workflows.
- How deep it goes
- Explains the conversational sequence and graph routing, but stops short of detailing individual tool implementations.
- Examples
- Uses order tracking, user identification, order lookup, cancellation, tracking numbers, and a LangGraph tools node.
- What you can do after
- Apply it via LangGraph to build an agent that selects tools, gathers missing parameters, and processes multi-step requests.
Read how Build Python Web Apps with Streamlit explains tool usage →
Investing for Programmers
A tool use pattern is an agent architecture in which an agent delegates tasks to external tools, libraries, or APIs and integrates their results.
- How it is used
- Expands an agent’s functional range by delegating data retrieval, transformation, and computation to specialized systems.
- How deep it goes
- Defines the delegation pattern and outlines tool selection, call orchestration, and result integration without implementation details.
- Examples
- Uses a stock price API for real-time data and a sentiment analysis tool for market news.
- What you can do after
- Apply it by orchestrating specialized tools and integrating their outputs into an agent’s decisions.
Spring AI in Action
A tool call is a model-generated request for an application to invoke an external callable capability.
- How it is used
- Enables an LLM to obtain external data and perform application actions during conversational generation.
- How deep it goes
- Traces tool schemas, generated arguments, automatic invocation, tool-result messages, and final completion responses.
- Examples
- Demonstrates getCurrentTime answering a time-zone question through generated JSON arguments and a returned timestamp.
- What you can do after
- Operate tool interactions through Spring AI’s consistent provider abstraction and interpret the resulting conversation messages.
Building Reliable AI Systems
An AI capability in which a model invokes external tools to obtain information or perform actions.
- How it is used
- Grounds conversational answers in external data and enables agents to act through APIs and services.
- How deep it goes
- Works through function definition, JSON schema, API invocation, tool-result handling, and final response; sketches broader agent safeguards rather than implementing them.
- Examples
- Uses a Python weather assistant, Seattle and New York queries, and specialized travel-retrieval and weather tools.
- What you can do after
- Build it from scratch: connect a conversational assistant to a weather API through structured function calls and returned tool results.
Authorization in Action
A tool invocation is a request by an agent or software system to execute an operation provided by an external tool.
- How it is used
- Enables agents to access documents, databases, APIs, and workflows while pursuing tasks.
- How deep it goes
- Defines invocation and structured integration, but defers the authorization decision governing whether a call is allowed.
- Examples
- Uses document search, database queries, calendar access, internal APIs, and automated workflows as tool examples.
- What you can do after
- Operate structured tool integrations by discovering capabilities and invoking them with defined inputs.
AI Agents in Action, Second Edition
A tool call is a structured request from an AI system to invoke a tool and obtain its result.
- How it is used
- Uses executable tools and external data sources to extend agents beyond direct language-model responses.
- How deep it goes
- Explains tool calls as workflow tasks and demonstrates registration and tracing, but defers implementation details of tool internals.
- Examples
- Demonstrates an internal research-sources tool, a search workflow, function registration, and trace inspection.
- What you can do after
- Apply it via OpenAI Agents by registering a function, prompting its use, and verifying the resulting tool call in traces.
Read how AI Agents in Action, Second Edition explains tool usage →
AI Agents in Action
Function calling is an LLM interface capability that produces a structured request to invoke a named function with extracted arguments.
- How it is used
- Connects LLM requests to executable application functions so agents can take actions.
- How deep it goes
- Explains schemas, selection, argument extraction, execution, and parallel calls, while deferring ordered action sequences.
- Examples
- Uses recommendation and weather functions, JSON argument objects, generated tool specifications, and an agent execution loop.
- What you can do after
- Apply it via an LLM API by defining tools, interpreting structured calls, executing functions, and returning results.
Related concepts
- OpenAI APIuses · The OpenAI API is a programming interface that enables software applications to send requests to OpenAI models and services and receive responses.
- tool descriptionuses · A tool description is a structured specification of a callable tool's purpose, capabilities, and invocation interface, including its expected inputs.
- command-line interfacecontrasts with · A command-line interface is a text-based interface through which users or programs interact with software or a computer by entering commands and receiving output.
- reliable agentused by · An AI agent that performs tool use and multi-step workflows while taking safe actions.
- control looppart of · A control loop is a repeated process that observes a system’s state, compares it with a desired state, and uses feedback to guide corrective action.
- tooluses · A tool is an instrument or software capability that enables a person or system to perform a task, obtain information, or interact with resources.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.