Tool recommendation system finds every needed tool for complex tasks
Tool Recommendation for Multi-Step Agents: Finding Every Tool a Task Needs
Machine LearningInformation Retrieval
Summary
Tasks that need multiple steps often fail if even one required tool is missing, but there are too many tools to list them all at once for an AI system. The authors built a system that sits between the AI agent and the tools, figuring out exactly which tools the task needs and making all of them available. This system works with hundreds more tools than typical setups and checks permissions carefully to keep users in control. Their tests show it finds the right tools much better than previous search methods, helping agents complete complex tasks efficiently and with lower costs.
What this means in practice
- •For software engineering teams: Enable AI-based automation tools to access all necessary APIs and services dynamically for complex workflows without missing steps.
- •For cloud platform providers: Improve cloud AI gateways to manage user consent and credentials while exposing many more integrated services safely and efficiently.$Commercial implications: Allows cloud providers to offer enhanced AI integration platforms that support broad toolsets with fine-grained user control.
Authors
Sam Partee, River Stevens, Thierry Damiba, Valerie Fanelle
Abstract
A multi-step agent task fails when a single tool it needs is missing, yet the tools an agent may need are too many to list: the roughly 8,200 tools on the platform we study have definitions that together exceed the context window of any model we evaluate. We present Tool Recommendation, a deployed system that gives an agent all of them, hundreds of times more than a typical MCP server exposes, through a fixed interface of three tools. It runs at the gateway between agent and tools, executing every call under the end user's own credentials and checking it against the user's consent for each service and the tools the agent may use. We evaluate it on both a public retrieval benchmark and billed end-to-end session cost. On ToolRet, Tool Recommendation places every tool a query needs in the top ten 57% more often than BM25, a lexical ranking that commercial tool search offers, and compares favorably with published systems up to nine times its size. In paired multi-step workflows, its first recommendation covered every required tool in 93% of Codex sessions, and against a fixed list of 100 tools it cut input tokens by 17% with Codex and 40% with Claude Code and the total bill in all 14 runs, by up to 40%, with no change in task completion.