tool call
In plain wordsA single request from an AI model to use one of its tools — naming the tool and filling in the details — which the surrounding software then carries out and reports back on.
A structured message in which a model requests one invocation of a tooltool, n.: Any outside capability an AI model is allowed to use — a web search, a calculator, a code runner, a database or an app — described to the model so it can ask for it by name., naming it and supplying arguments that match its schema, usually as JSON with a unique call identifier. The model cannot act by itself: the harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car. parses the call, checks permissionspermission system, n.: The rules that decide what an AI agent may do on its own, what it must ask a person about first, and what it may never do — like a phone's settings for which apps can use the camera., executes it and returns a tool result tagged with the same identifier, which joins the context for the model's next step. One model turn may contain several tool calls, run in sequence or in parallel. Tool calls are the atomic actions of the agent loopagent loop, n.: The repeating cycle at the heart of an AI agent: the model decides what to do next, software carries it out, the result is shown to the model, and the cycle repeats until the task is finished. and the unit in which agents are logged, audited, rate-limited and blocked — hookshook, n.: A user-supplied script or check that runs automatically at set moments in an AI agent's work — before a command, after a file edit, when it tries to stop — enforcing rules the AI cannot skip. and approval prompts typically act on individual calls. Malformed calls (invalid JSON, wrong argument types, invented tool names) are a classic failure, reduced by schema-constrained decoding and by error messages that let the model retry.
The transcript shows 38 tool calls, mostly file reads, before the agent made its first edit.
Origin
The phrase followed OpenAI's function callingfunction calling, n.: A feature of AI model services that lets the model reply with a neatly formatted request to run one of a program's functions, rather than with ordinary text; the program, not the model, then runs it. API (June 2023) and its later 'tools' interface, whose responses carry a 'tool_calls' field; Anthropic's API calls the same object a 'tool_use' block. By 2024 'tool call' was the generic name across providers and agent frameworksagent framework, n.: A software library that gives developers ready-made parts for building AI agents — model connections, tool handling, memory, multi-agent coordination and logging — so they need not write everything from scratch..
Usage note
Strictly, the call is the model's request and the tool result is the harness's reply; together they form one round trip of the agent loop.
Related
- tooltool, n.: Any outside capability an AI model is allowed to use — a web search, a calculator, a code runner, a database or an app — described to the model so it can ask for it by name.
- function callingfunction calling, n.: A feature of AI model services that lets the model reply with a neatly formatted request to run one of a program's functions, rather than with ordinary text; the program, not the model, then runs it.
- agent loopagent loop, n.: The repeating cycle at the heart of an AI agent: the model decides what to do next, software carries it out, the result is shown to the model, and the cycle repeats until the task is finished.
- harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car.
- hookhook, n.: A user-supplied script or check that runs automatically at set moments in an AI agent's work — before a command, after a file edit, when it tries to stop — enforcing rules the AI cannot skip.
- permission systempermission system, n.: The rules that decide what an AI agent may do on its own, what it must ask a person about first, and what it may never do — like a phone's settings for which apps can use the camera.
Sources
- Paniego, S. & Roy Gosthipaty, A. (2026). Harness, Scaffold, and the AI Agent Terms Worth Getting Right. Hugging Face blog. huggingface.co/blog/agent-glossary
Mentioned in
AI agentAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done., guardrailsguardrails, n.: Safety checks built around an AI system — filters on what goes in and comes out, and limits on what it may do — to keep it from producing harmful or off-topic content or leaking data., model APImodel API, n.: A web service through which programs send requests to an AI model run by a company and receive its answers, usually paying per use instead of running the model themselves., actionaction, n.: A choice an AI agent makes at one moment — move left, place a stone, type a command — which changes its situation and may earn it a reward., actuatoractuator, n.: The 'muscle' of a machine: a motor or similar device that turns electrical, hydraulic or air power into movement, such as turning a robot's joint or closing its gripper., agent loopagent loop, n.: The repeating cycle at the heart of an AI agent: the model decides what to do next, software carries it out, the result is shown to the model, and the cycle repeats until the task is finished., chat modelchat model, n.: A language model trained to hold a back-and-forth conversation, keeping track of who said what and replying as a helpful assistant. It is the engine inside chat apps., code executioncode execution, n.: Letting an AI write a small program and actually run it — to do math, analyze a spreadsheet or test its own code — then use the real output instead of guessing., compactioncompaction, n.: Like replacing a stack of rough notes with a one-page digest: when an AI's conversation grows too long for its memory, the history is condensed into a summary and the work continues from there., function callingfunction calling, n.: A feature of AI model services that lets the model reply with a neatly formatted request to run one of a program's functions, rather than with ordinary text; the program, not the model, then runs it., harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car., indirect prompt injectionindirect prompt injection, n.: A prompt-injection attack in which the trick instructions are hidden in content the AI reads on its own — a web page, an email, a file — rather than typed by the person using it., latencylatency, n.: How long you wait between asking an AI system for something and getting the answer. Lower latency means a snappier, more responsive system., permission systempermission system, n.: The rules that decide what an AI agent may do on its own, what it must ask a person about first, and what it may never do — like a phone's settings for which apps can use the camera., special tokenspecial token, n.: A reserved symbol in an AI model's vocabulary that is not ordinary text but a signal — 'start of message', 'end of answer', 'begin thinking' — used to structure what the model reads and writes., streamingstreaming, n.: Sending an AI's answer piece by piece while it is still being written, so the words appear on screen one after another instead of all at once at the end., structured outputsstructured outputs, n.: Getting an AI model to answer in an exact, machine-readable format — like a form with fixed fields — so other software can use the answer without guesswork., tooltool, n.: Any outside capability an AI model is allowed to use — a web search, a calculator, a code runner, a database or an app — described to the model so it can ask for it by name., tool usetool use, n.: An AI model's ability to reach beyond its own knowledge by operating outside tools — searching the web, running code, using a calculator or an app — and folding the results into its work., trajectorytrajectory, n.: The record of what happened as an AI agent acted: the sequence of situations it was in, the actions it took and the rewards it received., AgentDojoAgentDojo, prop. n.: A testing environment that measures how easily AI assistants that use email, calendars or banking tools can be hijacked by hidden instructions planted in the content they read., agentic reinforcement learningagentic reinforcement learning, n.: Training an AI assistant by rewarding success on long, multi-step jobs where it must use tools, browse, run code and react to what happens, rather than answer in one go., Claude Plays PokémonClaude Plays Pokémon, prop. n.: A 2025 livestream in which Anthropic's Claude AI played the classic Game Boy game Pokémon Red by itself, watched by thousands as it struggled and slowly progressed; it inspired similar runs by other models., code modecode mode, n.: A way for AI agents to use tools by writing a short program that calls them, instead of asking for each tool one step at a time.