AI 中文总结
本研究通过概率抽样揭示MCP注册表真实状态：仅48.8%服务器可启动，硬一致性高但安全注释缺失严重，并发现BFCL基准存在大量重复，而真实工具重复率极低。
AI 中文摘要
对模型上下文协议（MCP）服务器生态系统的研究，其样本抽取方式会悄然偏向于选择那些可正常工作的服务器：参考集、流行度列表、手工策划的框架，或修复服务器直至其启动的流水线。我们报告了未经修复的概率样本实际包含的内容。从包含24,135个服务器的注册表普查中，我们抽取了400个带有已发布种子的npm/stdio服务器，并通过网络对每个服务器进行探测。仅48.8%的服务器完成了初始化握手，而使用相同工具测量的手工策划框架的完成率为66.7%，主要失败原因并非缺少凭据（13.3%），而是服务器根本无法启动（37.5%）。在195个实际运行的服务器中，硬一致性是全面的：在2,766个广告工具中，JSON Schema致命违规为零。可选的 safety 注释才是真正的差异来源，随机抽取中工具级遗漏率为58.8%，而策划框架中为41.5%，因此策划也美化了这一数字。然后，我们使用一种保持不变的方法，将这些服务器广告的工具描述与两个工具使用基准语料库进行比较。真实MCP工具在余弦相似度0.70时显示出2.8%的近似重复，且所有这些重复都发生在单个服务器内部：在测试的所有阈值下，跨作者的近似重复率为0.0%。BFCL v4显示16.7%的近似重复，其中16.4个百分点发生在独立呈现的任务之间。UltraTool显示0.3%，比真实工具更干净，因此这是BFCL的特性，而非合成语料库的类别特性。另外，原始BFCL行中68.8%和原始UltraTool行中85.6%是名称加描述的精确重复，而真实MCP中为0.4%，因此，在没有全局去重的情况下，对这些发布版本计算的任何统计量衡量的是重复而非工具。所有数据均可从已发布的脚本和已发布的种子重新生成。
英文摘要
Studies of the Model Context Protocol (MCP) server ecosystem draw their samples in ways that quietly select for servers that work: reference sets, popularity lists, hand-curated frames, or pipelines that repair a server until it starts. We report what an unrepaired probability sample actually contains. From a 24,135-server registry census we draw 400 npm/stdio servers with a published seed and probe each one over the wire. Only 48.8% complete an initialize handshake, against 66.7% for a hand-curated frame measured with the same instrument, and the dominant failure is not missing credentials (13.3%) but servers that never start at all (37.5%). Among the 195 that do run, hard conformance is total: zero fatal JSON Schema violations across 2,766 advertised tools. Optional safety annotations are the real variance, and the tool-level omission rate on a random draw is 58.8% against 41.5% on the curated frame, so curation flatters this figure too. We then compare the tool descriptions these servers advertise against two tool-use benchmark corpora using one method held constant. Real MCP tools show 2.8% near-duplication at cosine 0.70, and all of it lies within single servers: cross-author near-duplication is 0.0% at every threshold tested. BFCL v4 shows 16.7%, of which 16.4 points lie between independently presented tasks. UltraTool shows 0.3%, cleaner than real tools, so this is a property of BFCL and not of synthetic corpora as a class. Separately, 68.8% of raw BFCL rows and 85.6% of raw UltraTool rows are exact name-plus-description repeats, against 0.4% for real MCP, so any statistic computed over these releases without global deduplication measures repetition rather than tools. All figures regenerate from released scripts and a published seed.
Comments10 pages, 3 figures. Seeded, re-runnable pipeline and per-server outcomes: https://github.com/itguruhaseeb/mcp-probe ; archived at doi:10.5281/zenodo.21347997
