Data Profiler MCP runs on your own machine — the client starts it, so there is no endpoint to ping. 168 installs a week from pypi. Last commit 22 Aug 2026.
Profiles CSV/Parquet/Excel/JSON files: schema, stats, quality flags and dtype tips for LLM agents.
Today is the operative word: we check Data Profiler MCP every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add data-profiler-mcp -- uvx data-profiler-mcp
{
"mcpServers": {
"data-profiler-mcp": {
"args": [
"data-profiler-mcp"
],
"command": "uvx"
}
}
}
[mcp_servers.data-profiler-mcp]
command = "uvx"
args = ["data-profiler-mcp"]
{
"mcpServers": {
"data-profiler-mcp": {
"args": [
"data-profiler-mcp"
],
"command": "uvx"
}
}
}
{
"mcpServers": {
"data-profiler-mcp": {
"args": [
"data-profiler-mcp"
],
"command": "uvx"
}
}
}
Profile local CSV, Parquet, JSON and Excel files into a compact data-quality summary.
Tabular data for AI agents — load CSV/JSON/XLSX, query, group-by, stats (no database).
Read-only WHOOP v2 data for MCP clients, with local SQLite cache and CSV/JSONL/Parquet exports.
Semantic, format-aware diff of config files (JSON/YAML/TOML/INI/.env/XML/CSV) for AI agents
Local CSV, JSON and Excel conversion. Files are never uploaded and never enter the context.
Query a folder of CSV / Parquet / JSON files with SQL — one queryable source, read-only.
Create interactive visualizations and query data sources (SQLite, CSV, Parquet, JSON)
Contract testing and quality gates for spreadsheets created by AI agents.
Answers built from our own checks of this server.