The specific mechanism and moment where an AI agent decides which available tool or function to call next during a task — a distinct, often-debated design decision in agent engineering, separate from the reasoning that leads up to it.
REAL-WORLD EXAMPLE
"The whole run failed because of one bad agent tool choice three steps in."
LORE & ORIGIN
Became an increasingly explicit part of agent-engineering vocabulary as frameworks exposed tool selection as its own configurable, inspectable decision point (mirroring the literal tool_choice parameter exposed by several LLM APIs) — engineers now discuss, tune, and debug tool choice as its own discipline, separate from prompt design or the harness surrounding the agent, since a wrong tool choice can derail an otherwise well-reasoned task.
MENTIONS OVER TIME
Radar observation data not yet available for this term.
RELATED SEARCHES
TOP PLATFORMS
Not enough Radar data yet to show a platform breakdown.
WHEN DID YOU FIRST HEAR THIS?
CLICK AN OPTION BELOW TO CAST YOUR VOTE.
[ VOTE TO REVEAL COMMUNITY RESULTS ]