Presents a component-centric PoC dataset of malicious MCP servers and a two-stage behavioral deviation detector Connor achieving 94.6% F1-score.
A measurement study of model context protocol ecosystem
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 5verdicts
UNVERDICTED 5roles
background 1polarities
background 1representative citing papers
Catalogues five MCP server architectural patterns observed across fifteen servers, plus anti-patterns and quantitative measurements on tool handling and overhead.
BIV audits AI agent skills at scale, finding 80% deviate from declared behavior on 49,943 skills and achieving 0.946 F1 for malicious skill detection.
VIPER-MCP detects and exploits taint-style vulnerabilities in Model Context Protocol servers via anchor-query static analysis and feedback-driven prompt evolution, uncovering 106 zero-day vulnerabilities across 39,884 repositories with 67 CVEs assigned.
Introduces Task2MCP dataset and T2MRec model for recommending MCP servers to LLM agents based on task semantics and engineering constraints.
citing papers explorer
-
Behavioral Integrity Verification for AI Agent Skills
BIV audits AI agent skills at scale, finding 80% deviate from declared behavior on 49,943 skills and achieving 0.946 F1 for malicious skill detection.