{"slug": "cartograph-federated-tool-discovery-with-operator-attested-retrieval-for-ai", "title": "Cartograph: Federated Tool Discovery with Operator-Attested Retrieval for AI Agents", "summary": "Cartograph, a federated Model Context Protocol (MCP) proxy, reduces agent-visible tool discovery from O(n) catalog traversal to O(k) progressive disclosure, exposing three proxy tools instead of 374 definitions on a 22-server, 374-tool deployment, according to an arXiv paper (2609.30293v1). A 49-query author-constructed benchmark yields R@5 of 0.816 versus 0.592 for a Jaccard keyword baseline, while a measured top-5 discovery exchange uses 475 tokens instead of 42,450 under full-catalog accounting. The system combines Ed25519-signed operator-attested capability cards, a three-layer confusable-cluster analysis called Rift that identified 49 confusable clusters including four HIGH-risk clusters, and two-stage retrieval that ranks servers before tools, adding 5ms mean latency (0.8%) over direct stdio MCP calls across ten trials.", "body_md": "arXiv:2609.30293v1 Announce Type: new \nAbstract: The Model Context Protocol (MCP) enables AI agents to discover and call tools, but loading every definition becomes expensive as connected catalogs grow. We present Cartograph, a federated MCP proxy that changes agent-visible tool discovery from $O(n)$ catalog traversal to $O(k)$ progressive disclosure. Cartograph combines three mechanisms: (1) operator-attested capability cards, Ed25519-signed descriptions generated under the deploying operator's control rather than ranked publisher copy; (2) Rift, a three-layer confusable-cluster analysis comprising density clustering, query-margin analysis, and token diagnosis; and (3) two-stage retrieval, which ranks servers before tools. On a 22-server, 374-tool deployment, Cartograph exposes three proxy tools instead of 374 definitions. A 49-query author-constructed benchmark yields R@5 of 0.816, compared with 0.592 for a Jaccard keyword baseline, while a measured top-5 discovery exchange uses 475 tokens rather than 42,450 under the stated full-catalog accounting. Rift identifies 49 confusable clusters, including four HIGH-risk clusters in bootstrap-generated cards. An exploratory comparison of 119 LLM-generated descriptions removes the observed zero-distance cluster but shows that mixing card-generation regimes can reduce R@5. Gateway measurements over ten trials add 5ms mean latency (0.8%) relative to direct stdio MCP calls. Cartograph is complementary to code-execution approaches: it controls which tool descriptions are surfaced and records the provenance of the descriptions used for ranking for each query.", "url": "https://wpnews.pro/news/cartograph-federated-tool-discovery-with-operator-attested-retrieval-for-ai", "canonical_source": "https://arxiv.org/abs/2609.30293", "published_at": "2026-09-28 04:00:00+00:00", "updated_at": "2026-09-28 04:18:21.342092+00:00", "lang": "en", "topics": ["ai-agents", "agent-protocols", "ai-tools", "ai-research", "large-language-models"], "entities": ["Cartograph", "Model Context Protocol", "Rift", "arXiv", "Ed25519"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/cartograph-federated-tool-discovery-with-operator-attested-retrieval-for-ai", "markdown": "https://wpnews.pro/news/cartograph-federated-tool-discovery-with-operator-attested-retrieval-for-ai.md", "text": "https://wpnews.pro/news/cartograph-federated-tool-discovery-with-operator-attested-retrieval-for-ai.txt", "jsonld": "https://wpnews.pro/news/cartograph-federated-tool-discovery-with-operator-attested-retrieval-for-ai.jsonld"}}