ToolDNS slashes AI agent search space by 95% using DNS
Researchers retrofit semantic tool discovery onto the internet's most resilient substrate.
The coming era of autonomous AI agents demands scalable discovery across millions of tools, but existing solutions suffer from O(N) complexity and centralized governance. In a new arXiv paper (2607.18242), researchers Enhao Chen and Yulin Shao propose ToolDNS, a radical framework that embeds functional intent and organizational trust into the Domain Name System's hierarchical namespace. This transforms expensive semantic search into lightweight, O(log N) name resolution via three protocol-compliant enhancements: partially unfolded names, EDNS0 intent payloads, and logical subdomains.
To validate ToolDNS across the fragmented tooling landscape, the team constructed a large-scale heterogeneous benchmark comprising 33,688 real-world tools spanning MCP, A2A, RESTful, and Skill protocols. Results show ToolDNS slashes the per-query search space by 95.26% while matching state-of-the-art retrieval accuracy. Its UDP-native design reduces discovery latency by orders of magnitude compared to HTTP-based registries. The work demonstrates that scalable AI interoperability requires not more middleware, but smarter use of existing infrastructure.
- ToolDNS leverages DNS hierarchical namespace for O(log N) instead of O(N) tool search.
- Achieves 95.26% search space reduction on a benchmark of 33,688 tools across 4 protocols.
- UDP-native design cuts latency by orders of magnitude versus HTTP-based registries.
Why It Matters
Enables scalable, decentralized tool discovery for autonomous AI agents without adding fragile new infrastructure.