관심도 점수
100
- 검색 slug
- mcp
- 실제 연결
- Hacker News, GitHub, arXiv
- 키워드
- the, and, mcp, for, with, model, tool, that
- 마지막 수집
- 2026-09-21T10:36:22.077Z
출처별 탭
MCP was always a bad idea?
Measure contraction property on isometric leaves and monotone fibres
For finite measures with positive densities on convex Euclidean supports, we prove that $MCP(κ,N)$ passes with unchanged parameters to almost every isometric leaf of an arbitrary nonexpansive map. The proof rests on a sharp contraction inequality for geometric conditional densities, with exponent equal to the leaf codimension. The inherited dimension parameter is optimal. A total-variation limit on resolvent graphs extends the result to inverse fibres of maximal monotone relations, including convex gradient fibres...
Oracle high-dimensional $M$-estimation using smooth reparameterization for sparsity
This paper establishes a unified non-linear regularization framework for high-dimensional $M$-estimation, encompassing both linear models and Cox's proportional hazards models. Rather than relying on traditional additive non-convex penalties which pose severe optimization challenges, the proposed paradigm embeds sparsity directly into the transformation for the physical parameter $θ=φ^{(ν)}(β)$ using a smooth ($C^2$) component-wise "ReParametrization map for Sparsity (RePS)" $φ^{(ν)}$, and the penalty term is $λ\V...
Closed-World Resolution Against Tool Hallucination in LLM Agents
Tool-augmented large language model (LLM) agents fail in a way no tool-selection or tool-security method addresses: they call tools that do not exist and pass arguments no schema declares. Existing defenses either pick the right tool (selection) or constrain what an agent may do with real tools (gating), both of which presuppose the emitted call refers to a real tool at all. We show this is a structural blind spot: a hallucinated call is by construction not a decision any gate made, so no gate can reject it. This...
Characterizing Network Centralization and Observability in the Remote MCP Ecosystem
The Model Context Protocol (MCP) has emerged as the dominant interface for connecting autonomous agents to external data sources and execution environments. The ecosystem's transition from local process execution to remote Streamable HTTP deployments introduces unmeasured architectural and security constraints at scale. This paper presents a three-tier observability framework comprising catalog metadata (O_0), passive compliance signals (O_1), and live vulnerability analysis (O_2), applied to empirically character...
When Agents Look Like Beacons: NIDS Evasion by Model Context Protocol Traffic
The Model Context Protocol (MCP) standardizes communication between autonomous Artificial Intelligence (AI) agents and remote tools over Streamable HTTP. This shift introduces a class of machine-generated, authenticated, and high-frequency JSON-RPC traffic directly into enterprise networks. Enterprise network defenders have historically relied on machine-like cadence as an Indicator of Compromise (IoC). In this study, we show that without explicit network-layer indication, MCP traffic structurally and temporally r...
Position: It is Time to Virtualize Foundation Models with a Self-evolving Operating System Layer
AI applications have shifted from single, monolithic foundation models (FM) to compound agentic systems. Yet today's stacks remain fragmented: even as protocols (e.g., MCP, A2A) ease tool/agent connectivity, each framework embeds an implicit runtime for state, memory, budgets, and guardrails, making behavior non-portable and governance brittle. It mirrors computing before operating systems, when every program re-implemented basic services. This position paper argues that the field now needs a Foundation Model Oper...
Measuring and Exploiting Implicit Trust in LLM Tool-Calling Pipelines
The Model Context Protocol (MCP) enables LLMs to invoke external tools, but every tool interaction exposes the model to attacker-controlled text through multiple input channels (tool descriptions, tool results, sampling messages) that share a single context window without privilege separation. In this paper, we present a framework to measure the trust profile of an arbitrary LLM based on a variety of payload framings sent through different channels. Following this assessment, we devise cross-channel fragmentation...
PentestChain: A Cost-Aware, MCP-Orchestrated Framework for Automated Penetration Testing with Free-Tier LLMs
AI-driven penetration testing has been demonstrated with premium frontier models such as GPT-4, but the per-engagement token cost makes continuous, automated testing unaffordable for the smaller organisations that need it most. This paper presents PentestChain, a ten-phase automated penetration testing framework that couples a curated, deterministic exploit map with a cost-aware AI cascade-a local Ollama model (qwen2.5-7b) first, then free-tier OpenRouter and Cerebras, with a rule-based fallback that always produc...
TuiML: Machine Learning for AI Agents
Machine-learning libraries such as Weka and scikit-learn were designed for human programmers. Language-model agents now use these same libraries by recalling APIs from memory and writing code, an approach that hides what a library offers, delays errors until runtime, and loses experimental state between turns. We present TuiML, a self-contained machine-learning library built for AI agents, with native algorithms across supervised, unsupervised, time-series, data handling, tuning, and evaluation tasks. Every compon...
A reanalysis of the megamaser Hubble constant: from spot catalogues to peculiar velocities
The Hubble tension motivates careful scrutiny of the redshift-independent distances underpinning the local distance ladder. Water megamasers afford a purely geometric distance from orbital dynamics in an accretion disc. For five of the six Megamaser Cosmology Project (MCP) galaxies, we present a Bayesian forward model that infers the angular-diameter distance and warped Keplerian disc parameters jointly from the very long baseline interferometry quasi-observables: maser spot positions, velocities, and acceleration...
Graphify-Labs/graphify
Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.
Panniantong/Agent-Reach
Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
affaan-m/ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
koala73/worldmonitor
Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface
farion1231/cc-switch
A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io
punkpeye/awesome-mcp-servers
A collection of MCP servers.
open-webui/open-webui
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
langgenius/dify
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.
n8n-io/n8n
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
Snailclimb/JavaGuide
Java 面试 & 后端通用面试指南，覆盖计算机基础、数据库、分布式、高并发、系统设计与 AI 应用开发