Server definition

Hash: sha256:82cb99f45819f2229180a38432633c823b8baa13649bf7344c63b768d8f87479
What it is: What a remote MCP server returned when asked what it offers: 3 tools

The blob, as servednamed by its sha256

{
  "instructions": "Official Xinyuanyi MCP server: query signed AI-inference-storage benchmarks (R1-R9), search Xinyuanyi's technical articles, and estimate KV-cache tiering ROI. All numbers carry report citations; see https://mingxinstorage.xyz/en for sources.",
  "tools": [
    {
      "description": "Estimate the ROI of adding a Xinyuanyi FX100 KV-cache storage tier to a GPU inference cluster. Model is a faithful port of the reproducible Python model (accel_value.py). Results are mid-scenario estimates, not commitments.",
      "inputSchema": {
        "properties": {
          "arrays": {
            "description": "FX100 array count, default 8",
            "type": "number"
          },
          "cold_share": {
            "description": "Optional cold-recovery load share (0.1-0.5 estimated band)",
            "type": "number"
          },
          "gpus_per_node": {
            "description": "GPUs per node, default 8",
            "type": "number"
          },
          "nodes": {
            "description": "GPU node count, default 16",
            "type": "number"
          },
          "uplift": {
            "description": "Optional custom throughput uplift (0.29-0.40 measured band)",
            "type": "number"
          }
        },
        "type": "object"
      },
      "name": "estimate_roi",
      "outputSchema": null
    },
    {
      "description": "Query Xinyuanyi's signed benchmark results for FX-series storage acceleration: throughput +29-40%, TTFT -26-32% (480B model on 8x AMD MI308X), model loading 6.2-9.3x vs NFS, and the full R1-R9 report list with hosted PDF URLs. All numbers come from signed test reports; reproducible via github.com/mingxin-tech/mingxin-kvcache-bench.",
      "inputSchema": {
        "properties": {},
        "type": "object"
      },
      "name": "query_benchmark",
      "outputSchema": null
    },
    {
      "description": "Search Xinyuanyi's published articles on AI inference storage (KV cache tiering, NVMe-oF all-flash arrays, LLM serving). Returns titles, URLs and excerpts. Supports Chinese and English.",
      "inputSchema": {
        "properties": {
          "lang": {
            "description": "Language filter, default en",
            "enum": [
              "zh",
              "en"
            ],
            "type": "string"
          },
          "limit": {
            "description": "Max results (1-10), default 5",
            "type": "number"
          },
          "query": {
            "description": "Search keywords (Chinese or English)",
            "type": "string"
          }
        },
        "required": [
          "query"
        ],
        "type": "object"
      },
      "name": "search_mingxin_docs",
      "outputSchema": null
    }
  ]
}

Verify it yourself

curl -s https://api.teppi.xyz/v1/evidence/sha256:82cb99f45819f2229180a38432633c823b8baa13649bf7344c63b768d8f87479 | sha256sum