GetQueryly: MCP Server and API for AI Data Analysis (CSV to Charts in Minutes)

GetQueryly has an MCP server and API for AI data analysis. Upload a CSV, Excel, JSON, PDF, or Parquet file, ask in plain English, and get charts with short explanations plus a Quick Insight summary. It works out of the box with Claude, Cursor, and any MCP client, or via straight REST from your own app.

• Docs: https://getqueryly.com/api-docs
• Product: https://getqueryly.com

Why an MCP Server for Data Analysis?

AI assistants often guess about your data because they cannot see it directly. Point GetQueryly at your file, and the guesswork stops. The assistant calls a tool, the calculations run on your actual upload, and every result ships with a chart that proves it.

Use it when you want average revenue by region with visual proof, outliers flagged before you chart, or a consensus check from three distinct analyst passes on a hard call.

Setup: MCP in 5 Minutes

Endpoint: https://getqueryly.com/mcp/sse

1. Create an account at getqueryly.com.
2. Open Dashboard > API Keys > Generate key.
3. Ensure MCP is enabled on that key (toggled per key).
4. Add the following to your MCP client config:

{
  "mcpServers": {
    "getqueryly": {
      "url": "https://getqueryly.com/mcp/sse",
      "headers": {
        "X-API-Key": "dmk_your_key_here"
      }
    }
  }
}

5. Restart your client and ask "list getqueryly tools" to verify the connection.

Setup: REST API for Servers & Scripts

Same key, same costs. Send your key via the X-API-Key header.

Base URL: https://getqueryly.com

Upload a File

Max file size: 50MB. Supported types: CSV, XLSX, XLS, JSON, PDF, TXT, MD, PARQUET.

curl -X POST https://getqueryly.com/api/data-scientist/upload \
  -F "session_id=ds_1720000000_abc12" \
  -F "file=@sales_data.csv"

Query the Session

curl -X POST https://getqueryly.com/api/data-scientist/query \
  -F "session_id=ds_1720000000_abc12" \
  -F "query=Show average revenue by region"

Health Checks, Insights & Chat

Health check

curl -H "X-API-Key: dmk_your_key_here" \
  https://getqueryly.com/api/data-scientist/health/ds_1720000000_abc12

Insights

curl -H "X-API-Key: dmk_your_key_here" \
  https://getqueryly.com/api/data-scientist/insights/ds_1720000000_abc12

Chat

curl -X POST https://getqueryly.com/api/chat \
  -H "X-API-Key: dmk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"message":"Summarize this session for standup"}'

Check Credit Limits ("Makes")

curl -H "X-API-Key: dmk_your_key_here" \
  https://getqueryly.com/api/makes/limits

Tool Costs & Pricing Model

• Upload: Free
• Chat: 1 Make
• Data Query: 2 Makes
• Health Check: 1 Make
• Insights: 2 Makes
• Swarm Quick: 4 Makes
• Swarm Full: 6 Makes
• Report: 1 Make

Pro-Tip: Upload once and reuse the same session_id for every follow-up to save credits. Check your limits before running a swarm analysis. Failed runs auto-refund so your balance is protected. Free tier covers daily testing, and paid packs stack.

3 Workflows Worth Copying

Health Check Before Charts

• Prompt: "Upload ./sales.csv with GetQueryly, run ds_health_check, and report missing values, outliers, and bad types."
• Result: You get a data quality score out of 100 alongside per-column missing rates. Never chart dirty data again.

Ask with Proof

• Prompt: "Run ds_query on that session for average revenue by region, output a bar chart, and include a one-line summary under each bar."
• Result: Returns precise numbers, visual charts, and ready-to-share summaries for Slack.

Consensus for Hard Calls

• Prompt: "Run ds_swarm_quick on churn drivers, then list where the analysts agree."
• Result: Synthesizes independent analytical passes to catch nuances a single pass might miss.

FAQ

Is GetQueryly an AI data analysis API?
Yes. Upload, query, check health, generate insights, chat, and export over REST using a single header. View full details in the API Docs.

Does it work as an MCP server for Claude and Cursor?
Yes. Add the SSE URL and your API key to your client config. Exposed tools include ds_upload, ds_query, ds_health_check, ds_insights, ds_swarm_quick, ds_swarm_full, ds_report, ai_chat, chart, and more.

What file formats are supported?
CSV, Excel (XLSX/XLS), JSON, PDF, TXT, Markdown, Parquet, and images containing text up to 50MB per upload over REST.

Can I share results without login credentials?
Yes. One click generates a public share link containing the charts and narrative. Viewers do not need an account, while raw file downloads remain owner-only.

Where do I start?
Head over to getqueryly.com, grab an API key, check the API Docs, and run your first upload curl. I built GetQueryly and will be hanging out in the comments! Post your file type and question below, and I'll reply with the exact prompt and cheapest tool chain to use.