ToolACE-MCP: Generalizing History-Aware Routing from MCP Tools to the Agent Web
This work addresses the problem of efficient tool navigation for agents in open collaborative networks, representing an incremental improvement with strong specific gains.
The paper tackles the scalability and generality bottlenecks in the Agent Web and Model Context Protocol (MCP) ecosystem by proposing ToolACE-MCP, a pipeline for training history-aware routers that enables precise navigation in large-scale tool ecosystems, achieving superior performance on real-world benchmarks like MCP-Universe and MCP-Mark.
With the rise of the Agent Web and Model Context Protocol (MCP), the agent ecosystem is evolving into an open collaborative network, exponentially increasing accessible tools. However, current architectures face severe scalability and generality bottlenecks. To address this, we propose ToolACE-MCP, a pipeline for training history-aware routers to empower precise navigation in large-scale ecosystems. By leveraging a dependency-rich candidate Graph to synthesize multi-turn trajectories, we effectively train routers with dynamic context understanding to create the plug-and-play Light Routing Agent. Experiments on the real-world benchmarks MCP-Universe and MCP-Mark demonstrate superior performance. Notably, ToolACE-MCP exhibits critical properties for the future Agent Web: it not only generalizes to multi-agent collaboration with minimal adaptation but also maintains exceptional robustness against noise and scales effectively to massive candidate spaces. These findings provide a strong empirical foundation for universal orchestration in open-ended ecosystems.