Tool Use
Tool use is the capability of an AI model to do more than write text: it can call external tools, such as searching a database, sending an email or running a calculation, by emitting a structured request that software executes for real.
No jargon, no math
In plain words, what is it?
A language model on its own can only predict text. Tool use gives it hands. The model is shown a list of available tools, each with a name, a description and the inputs it needs. Mid-answer, when it decides it needs one (the stock level, the weather, a lookup in your CRM), it pauses and emits a structured call: tool name plus arguments, written in JSON. Your software executes the real action and hands the result back. The model reads the result and continues the answer with actual data in hand. Where it breaks: the model chooses the tool and the arguments, so a vague description or a wrong argument produces a wrong call. Clear tool descriptions, permission checks on execution and full logging are what keep this safe.
- 01
Ask. You ask something the model cannot know from text alone.
- 02
Choose. The model picks the right tool from the list it was given.
- 03
Call. It emits a structured request: tool name and arguments, nothing vague.
- 04
Execute. Your software runs the real action and returns the outcome.
- 05
Answer. The model continues with the actual data now in its context.
Terms that live nearby.
Function Calling
Function calling is the structured form of tool use: the model emits a machine-readable request naming a function and its arguments, written in JSON, which your code validates and executes, then feeds the return value back into the conversation.
ReadModel Context Protocol
Model Context Protocol (MCP) is an open standard that lets an AI agent connect to your tools and data (files, databases, CRMs, APIs) through one common plug, instead of a custom integration for every tool.
ReadStructured Output
Structured output is a mode where the model is forced to answer in a machine-readable format (JSON, a table, a fixed schema) instead of free text, so its answer can be validated, stored and passed to other software without a human retyping it.
Read§ 04
Want this running in your stack?
We design and ship AI systems with permissions, logs and human approval. It starts with a 30-minute scoping call.