AI News selected for Professionals and Decision Makers
Primary Research Stream

AI Tool Discovery at Scale: All You Need is DNS

06:00 · July 22, 2026 · arXiv cs.AI RSS

AI Tool Discovery at Scale: All You Need is DNS

The coming era of autonomous AI agents demands a discovery mechanism capable of navigating millions of tools, yet existing solutions buckle under O(N) complexity and centralized governance. Instead of building another fragile overlay, we propose ToolDNS, a radical framework that retrofits semantic tool discovery onto the Internet's most resilient substrate: the Domain Name System (DNS). By embedding functional intent and organizational trust into a hierarchical namespace, ToolDNS transforms an expensive semantic search into a series of lightweight, O(log N) name resolutions. We introduce three protocol-compliant enhancements to enable decentralized governance and semantic pruning: partially unfolded names, EDNS0 intent payloads, and logical subdomains. To rigorously evaluate this approach across the fragmented tooling landscape, we construct and release a large-scale heterogeneous benchmark comprising 33,688 real-world tools spanning MCP, A2A, RESTful, and Skill protocols. On this dataset, ToolDNS slashes the per-query search space by 95.26% while matching state-of-the-art retrieval accuracy. Furthermore, its UDP-native design reduces discovery latency by orders of magnitude compared to HTTP-based registries. Our work demonstrates that scalable AI interoperability requires not more middleware, but a smarter utilization of the infrastructure already beneath our feet.

Summary

The rapid growth of autonomous AI agents has exposed a core bottleneck in service discovery: locating the right tool among potentially millions of candidates without incurring prohibitive latency or relying on fragile centralized indexes. Existing approaches either perform exhaustive vector similarity searches over global registries or inject tool descriptions directly into large language model contexts, both of which scale poorly and introduce single points of failure or governance disputes across organizations.

ToolDNS addresses this by repurposing the Domain Name System itself as the discovery substrate. Functional intent, organizational trust, and protocol metadata are encoded directly into a hierarchical namespace under a proposed .tools top-level domain. Agents express queries through standard recursive resolution rather than separate search infrastructure, converting what would be linear scans into a sequence of lightweight name lookups whose cost grows logarithmically with the number of tools.

Three protocol-compliant extensions make this possible. Partially unfolded domain names allow progressive narrowing of an initially vague intent. EDNS0 extension payloads carry semantic descriptions without violating label-length limits. Logical subdomains decouple functional hierarchy from administrative control, so multiple independent entities can manage tools under the same parent branch while retaining their own governance and trust boundaries. The resulting system remains fully compatible with existing DNS resolvers and requires no new servers or client SDKs.

Evaluation on a newly released benchmark of 33,688 real-world tools spanning MCP, A2A, RESTful, and Skill protocols shows that two layers of hierarchical pruning shrink the per-query search space by 95.26 percent while preserving retrieval accuracy comparable to state-of-the-art vector methods. Because queries travel over UDP, measured discovery latency drops by orders of magnitude relative to HTTP-based registries. The framework and dataset are publicly available, offering a concrete path toward decentralized, infrastructure-native tool discovery that avoids additional middleware layers.

Why it matters

This research is highly relevant for Dutch AI infrastructure developers and researchers building multi-agent systems. Its decentralized governance model aligns well with European data sovereignty and transparent AI goals, offering a scalable alternative to centralized tool registries.

More in this beat
agent-skillsai-agentsevaluation-benchmarksllm-agentsmodel-context-protocolToolDNStool-use
SAAG: Structured Agent Assessment and Grounding

06:00 · July 22, 2026

SAAG: Structured Agent Assessment and Grounding

This research provides a rigorous framework for diagnosing and mitigating hallucinations in AI agents, directly supporting the Dutch and EU focus on transparent and trustworthy AI. It offers researchers new methodologies to evaluate agentic systems beyond simple binary exact-match metrics.

Relevance 85 · Audience 95

MedCalc-Pro: Solving Complex Medical Calculations with LLM Agents

06:00 · July 7, 2026

MedCalc-Pro: Solving Complex Medical Calculations with LLM Agents

This research is highly relevant for Dutch AI researchers and health-tech enterprises focusing on clinical decision support systems. The proposed benchmark and agent framework align with the Netherlands' strong emphasis on robust, validated, and ethical AI applications in healthcare.

Relevance 85 · Audience 95

Code execution with MCP: Building more efficient agents

01:00 · November 4, 2025

Code execution with MCP: Building more efficient agents

Highly actionable for Product Teams and Builders with concrete implementation patterns, code snippets, and measurable efficiency gains (e.g., 98.7% token reduction). Directly addresses model/product updates in agent tooling and context management.

Relevance 85 · Audience 90

Harnessing agent memory to build lifelong AI partners for materials scientists

06:00 · August 13, 2026

Harnessing agent memory to build lifelong AI partners for materials scientists

This research is highly relevant for Dutch AI researchers and high-tech materials enterprises looking to deploy autonomous AI agents for R&D. The proposed model-agnostic memory framework addresses critical challenges in AI reproducibility and workflow efficiency, offering actionable methodologies for advanced scientific computing.

Relevance 85 · Audience 95

SkillTrace: Multi-Trace Provenance Auditing for LLM-Agent Skill Reuse

06:00 · August 7, 2026

SkillTrace: Multi-Trace Provenance Auditing for LLM-Agent Skill Reuse

This research is highly relevant for Dutch AI researchers and enterprises focused on AI governance, IP protection, and compliance with EU transparency regulations. It provides a rigorous, actionable methodology for auditing LLM-agent ecosystems, which is crucial for maintaining ethical and transparent AI marketplaces.

Relevance 85 · Audience 95

AgentStream: How Well Do Self-Evolving LLM Agents Perform Under Streaming Tasks?

06:00 · August 4, 2026

AgentStream: How Well Do Self-Evolving LLM Agents Perform Under Streaming Tasks?

This research is highly relevant for Dutch AI researchers developing autonomous LLM agents, providing a rigorous framework for evaluating continuous learning in realistic deployment scenarios. Understanding how model capabilities gate self-evolution is crucial for building robust and reliable AI systems.

Relevance 85 · Audience 95

Calibrated Selective Fact-Checking via Evidence Chain Evaluation

06:00 · July 22, 2026

Calibrated Selective Fact-Checking via Evidence Chain Evaluation

This research is highly relevant for Dutch AI researchers and practitioners focusing on trustworthy and ethical AI, a key priority in the Netherlands and the EU. The abstention mechanism directly addresses LLM hallucination and reliability issues, offering actionable methodologies for building compliant, high-stakes verification pipelines under EU AI regulations.

Relevance 85 · Audience 95

Deterministic Replay for AI Agent Systems

06:00 · July 21, 2026

Deterministic Replay for AI Agent Systems

Directly actionable for Dutch AI researchers and advanced practitioners working on agent systems, offering high technical depth, reproducibility resources, and alignment with EU emphasis on transparent, reliable AI.

Relevance 85 · Audience 90