Convert PDF, DOCX, images, and HTML files to markdown via magicconvert-mcp-server (uvx)

{
  "server": "magicconvert v0.1.2",
  "package": "magicconvert-mcp-server",
  "launcher": "uvx magicconvert-mcp-server",
  "transport": "stdio",
  "framing": "NDJSON",
  "protocol_version": "2024-11-05",
  "tool_count": 4,
  "tools": ["convert_file_to_markdown", "convert_base64_file_to_markdown", "convert_url_to_markdown", "convert_text_to_markdown"],
  "trace": [
    {
      "id": 1,
      "method": "initialize",
      "result": {
        "serverInfo": {
          "name": "magicconvert",
          "version": "0.1.2"
        },
        "protocolVersion": "2024-11-05"
      }
    },
    {
      "id": 10,
      "tool": "convert_text_to_markdown",
      "args": {
        "text_content": "<h1>Agent Report</h1><p>The <strong>tani registry</strong> now has 72 exchange threads.</p>"
      },
      "result": {
        "content": [
          {
            "type": "text",
            "text": "<h1>Agent Report</h1><p>The <strong>tani registry</strong> now has 72 exchange threads.</p>"
          }
        ],
        "isError": false
      },
      "note": "Echoed HTML as-is — did NOT convert to markdown"
    },
    {
      "id": 11,
      "tool": "convert_file_to_markdown",
      "args": {
        "file_path": "/tmp/test-pathfinder.csv"
      },
      "result": {
        "content": [
          {
            "type": "text",
            "text": "name,role,language,experience_years
Alice,backend,Python,8
Bob,frontend,TypeScript,5
Charlie,devops,Go,12
Diana,fullstack,Rust,3
Eve,data,Python,6
"
          }
        ],
        "isError": false
      },
      "note": "CSV returned as raw text, not markdown table"
    },
    {
      "id": 12,
      "tool": "convert_file_to_markdown",
      "args": {
        "file_path": "/tmp/test-pathfinder.pdf"
      },
      "result": {
        "content": [
          {
            "type": "text",
            "text": "Hello from tani pathfinder!

"
          }
        ],
        "isError": false
      },
      "note": "PDF text extraction SUCCEEDED — correct content extracted"
    }
  ],
  "verified_at": "2026-06-11T02:16:00Z",
  "cold_start_ms": 4000,
  "tool_latency_ms": "~500-2000 per call"
}