Agent skill
processing-data
Processes CSV files and pandas DataFrames. Use when working with CSV files, tabular data, spreadsheets, or when the user asks to query, analyze, or manipulate structured data.
Install this agent skill to your Project
npx add-skill https://github.com/binome-dev/humcp/tree/main/src/tools/data
SKILL.md
Data Processing Tools
Tools for working with CSV files and pandas DataFrames.
CSV Operations
List available CSV files
result = await list_csv_files()
# Returns: {"success": True, "data": ["file1", "file2"], "count": 2}
Read CSV content
result = await read_csv_file("mydata", row_limit=100)
# Returns rows as list of dicts
Query with SQL (DuckDB)
result = await query_csv_file("mydata", "SELECT * FROM mydata WHERE value > 10")
Security: Only SELECT queries allowed. INSERT, UPDATE, DELETE rejected.
Add/remove CSV files
await add_csv_file("/path/to/file.csv")
await remove_csv_file("filename")
Pandas Operations
Create DataFrame
# From CSV
result = await create_pandas_dataframe(
dataframe_name="sales",
create_using_function="read_csv",
function_parameters={"filepath_or_buffer": "data.csv"}
)
# From dict
result = await create_pandas_dataframe(
dataframe_name="mydf",
create_using_function="DataFrame",
function_parameters={"data": {"col1": [1, 2], "col2": [3, 4]}}
)
Allowed functions: DataFrame, read_csv, read_json, read_excel, read_parquet, read_feather, read_orc, read_html, read_xml, read_table, read_sql.
Run operations
# Get first rows
result = await run_dataframe_operation("sales", "head", {"n": 5})
# Filter data
result = await run_dataframe_operation("sales", "query", {"expr": "amount > 100"})
# Get statistics
result = await run_dataframe_operation("sales", "describe", {})
Allowed operations: head, tail, describe, query, filter, groupby, sort_values, mean, sum, count, and 100+ more safe operations.
Export DataFrame
result = await export_dataframe(
dataframe_name="sales",
export_function="to_csv",
export_parameters={"path_or_buf": "output.csv", "index": False}
)
Manage DataFrames
await list_dataframes() # List all in memory
await get_dataframe_info("sales") # Get details
await delete_dataframe("sales") # Remove from memory
Security Notes
- CSV queries: Only SELECT statements allowed
- Pandas: Operations restricted to allowlist of safe methods
- No arbitrary code execution via getattr
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
binome-dev/humcp
llm
Generate text, call tools, and get structured JSON output using LLM providers (Claude, OpenAI, Gemini, Ollama). Use when the user needs to call an LLM API for text generation, function calling, or structured data extraction.
binome-dev/humcp
calendar-tools
Manages calendar bookings and video conferencing. Use when the user needs to check availability, create bookings on Cal.com, or schedule and manage Zoom meetings.
binome-dev/humcp
http-api-client
Make HTTP requests to any URL or API endpoint. Use when the user needs to call a REST API, fetch data from a URL, send webhooks, or test HTTP endpoints. Supports GET, POST, PUT, PATCH, DELETE methods with custom headers and JSON body.
binome-dev/humcp
persistent-memory
Store, search, and retrieve persistent memories and conversation history using Mem0 or Zep. Use when the user needs to remember facts, maintain context across sessions, or manage conversational memory.
binome-dev/humcp
storage
Use these tools for S3-compatible object storage operations with MinIO
binome-dev/humcp
social-media
Interact with X (Twitter) for searching tweets, posting tweets, and fetching user profiles and timelines.
Didn't find tool you were looking for?