A study analyzing the schemas of 4,951 public MCP servers revealed significant issues with tool descriptions and parameter clarity. Approximately 21.8% of parameters lacked any description, forcing models to guess their purpose. Furthermore, about one in six tools had descriptions that were not distinct from other tools on the same server, with 17.4% having descriptions that were less than 10% distinctive. This lack of clarity in tool and parameter descriptions hinders a model's ability to effectively interact with these servers. AI
IMPACT Lack of clear tool and parameter descriptions on MCP servers hinders AI model performance and increases the likelihood of errors.
RANK_REASON Analysis of public server schemas detailing tool and parameter descriptions. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →