reading-plus-ai/mcp-server-data-exploration
mcp-server-data-exploration is hosted online, so all tools can be tested directly either in theInspector tabor in theOnline Client.
If you are the rightful owner of mcp-server-data-exploration and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
MCP Server is a versatile tool designed for interactive data exploration.
Try mcp-server-data-exploration with chat:
Has a README
Github repo has a README.md.
Has a License
Github repo has license info.
Server can be inspected
View server inspector
Server schema can be extracted
Can get at lease one tool info from the README or server.
Online hosted on MCPHub
Can be automatically deployed by MCPHub.
Has social accounts
Do not have any social accounts.
Claimed by the author or certified by MCPHub
If you are the author, claim authorship
AI Evaluation ReportTotal Score: 9/10
The agent demonstrates strong capabilities in executing data exploration tasks, such as loading CSV files, cleaning headers, sorting data, and performing calculations. It successfully completed tasks involving data manipulation and provided accurate results. However, the agent occasionally includes unnecessary explanations that can detract from clarity and lacks detailed descriptions of intermediate steps in some responses. Overall, the agent performs well in executing the tasks but could improve by providing more concise and detailed explanations of the processes involved.
Test case 1
Score: 9/10Load /app/data/cities.csv into cities. Then use run_script to clean headers, sort by LatD descending, and print the top 5 rows (include city and state columns).It appears that the script to clean the headers and display the top 5 rows from the
citiesDataFrame did not return any output. However, I have confirmed the column names after cleaning, which are as follows:LatDLatMLatSNSLonDLonMLonSEWCityState
It seems like the column names for city and state have extra quotes. I will adjust the script to clean them again and attempt to print the top 5 rows. Let's try that. The script has successfully cleaned the headers and sorted the
citiesDataFrame by latitude in descending order. Here are the top 5 rows, including the city and state columns:City State Regina SA Winnipeg MB Vancouver BC Williston ND Tacoma WA Please let me know if you need any further analysis or modifications!
Test case 2
Score: 7/10Load /app/data/addresses.csv into addresses. Then use run_script to standardize column names (strip quotes/spaces, replace spaces with underscores), coerce the last column (postal code) to numeric, and print the average postal code (ignore non-numeric).The CSV file has been successfully loaded into the DataFrame named addresses. The average postal code calculated from the last column is 21769.8.
If you need any further analysis or tasks, feel free to ask!
Test case 3
Score: 10/10Load /app/data/crash_catalonia.csv into crashes. Then use run_script to clean headers, convert the 'Number of Crashes' column to numeric, and print total crashes and a table of crashes by Day of Week sorted descending.The data has been successfully processed. Here are the results:
Total Crashes
- Total Crashes: 118,489
Crashes by Day of Week
Day of Week Number of Crashes Friday 19,147 Thursday 17,954 Wednesday 17,394 Tuesday 17,337 Monday 17,279 Saturday 15,714 Sunday 13,664 The table shows the total number of crashes sorted in descending order by day of the week. If you need further analysis or information, feel free to ask!
