Data Visualization & Analysis Skill - User Guide

Enable your AI agents to analyze data and create visual charts from CSV, Excel, and other data formats using Python.

Overview

The Data Visualization & Analysis skill empowers your Swiftask AI agents to perform data analysis and generate visual charts directly within your conversations. This ready-to-use skill leverages Python's powerful data processing libraries to transform raw data into actionable insights through graphs, charts, and visualizations.

Key capabilities:

  • Analyze data from CSV, Excel, and other formats

  • Generate charts (bar, line, pie, scatter plots, etc.)

  • Perform statistical analysis

  • Create custom visualizations based on your requirements

  • Return high-quality chart images directly in chat


Prerequisites

Before using this feature, ensure you have:

  • An active Swiftask account with agent creation permissions

  • An existing AI agent or the ability to create a new one

  • Data files ready for analysis (CSV, Excel, or similar formats)

  • Basic understanding of your data structure


Step-by-Step Setup Guide

Step 1: Access Agent Configuration

  1. Navigate to your Swiftask workspace

  2. Open the Agents section from the main menu

  3. Select an existing agent or create a new one

  4. Click on the agent to open its configuration settings

Step 2: Add the Data Visualization Skill

  1. In the agent configuration panel, locate the Skills section

  2. Click on "Ready-to-Use Skills" or "Browse Skills"

  3. Use the search bar to find: "Data Visualization" or "Python Code Execution"

  4. Click "Add" or "+" button next to the skill

  5. Confirm the skill has been added to your agent's capabilities

Step 3: Save Configuration

  1. Review your agent's skills list to ensure the Data Visualization skill appears

  2. Click "Save" or "Update Agent" to apply changes

  3. Wait for confirmation that the agent has been updated successfully

Step 4: Start Using the Skill

  1. Create a new chat with your configured agent

  2. Upload your data file (CSV, Excel, etc.) or provide data inline

  3. Request a specific visualization or analysis

  4. The agent will process your data and return chart images


Practical Use Cases

Use Case 1: Sales Performance Analysis

Scenario: You have a CSV file with monthly sales data and want to visualize trends.

Request example:

Expected output: A line chart image displaying revenue progression over time with labeled axes and legend.


Use Case 2: Customer Demographics Breakdown

Scenario: Understand your customer base distribution by age group.

Request example:

Expected output: A pie chart with percentage breakdowns and color-coded segments.


Use Case 3: Comparative Analysis

Scenario: Compare performance metrics across different departments.

Request example:

Expected output: A grouped or stacked bar chart with clear department labels and metrics.


Use Case 4: Statistical Summary

Scenario: Get quick statistical insights from a dataset.

Request example:

Expected output: Text summary with statistics followed by a histogram visualization.


Best Practices & Tips

Data Preparation

  • Ensure your data files are clean and well-structured

  • Use clear column headers that describe the data

  • Remove unnecessary rows or columns before uploading

  • Check for missing values or inconsistencies

Effective Prompting

  • Be specific about the type of chart you want (line, bar, pie, scatter, etc.)

  • Mention which columns or fields to visualize

  • Specify any particular formatting preferences (colors, labels, titles)

  • Request multiple visualizations in separate messages for clarity

File Format Recommendations

  • CSV: Best for simple tabular data

  • Excel (.xlsx): Ideal for multi-sheet workbooks with complex data

  • Ensure files are not password-protected

  • Keep file sizes reasonable for faster processing

Optimization Tips

  • For large datasets, specify the exact data range you want to visualize

  • Request summary statistics before creating visualizations

  • Save frequently used chart configurations by noting successful prompts

  • Combine multiple related visualizations in a single analysis session


Troubleshooting

Problem: Agent doesn't recognize the data file

Solutions:

  • Verify the file format is supported (CSV, Excel)

  • Check that the file uploaded successfully

  • Try re-uploading the file

  • Ensure the file is not corrupted


Problem: Chart doesn't display as expected

Solutions:

  • Be more specific in your visualization request

  • Check that the column names in your request match those in the data

  • Request a data preview first to verify structure

  • Try a different chart type that better suits your data


Problem: "Error executing Python code" message

Solutions:

  • Simplify your request and try again

  • Check for special characters or formatting issues in your data

  • Ensure your data doesn't contain incompatible data types

  • Contact support if the error persists


Problem: Skill not appearing in agent configuration

Solutions:

  • Refresh the skills catalog

  • Check your account permissions

  • Verify your subscription plan includes custom skills

  • Clear browser cache and try again


Technical Details

Execution Environment

  • Python code runs in a secure E2B sandbox environment

  • Isolated execution ensures data security and system stability

  • No code persists after execution completes

Available Libraries

The skill includes access to popular Python data analysis libraries:

  • pandas: Data manipulation and analysis

  • matplotlib: Chart and graph creation

  • seaborn: Statistical data visualization

  • numpy: Numerical computing

Output Format

  • Charts are returned as PNG images

  • High-resolution output suitable for reports and presentations

  • Images are displayed directly in the chat interface


Frequently Asked Questions

Q: What file size limit applies to uploaded data? A: File size limits depend on your subscription plan. Generally, files up to 10MB are supported for most plans.

Q: Can the agent create multiple charts from one dataset? A: Yes, you can request multiple visualizations in sequence or ask for a comprehensive analysis with several charts.

Q: Is my data secure when using this skill? A: Yes, all data processing occurs in isolated sandbox environments and is not stored after execution.

Q: Can I customize chart colors and styles? A: Yes, include specific styling preferences in your request (e.g., "use blue and green colors" or "create a minimalist chart").

Q: Does this work with real-time data sources? A: When combined with Data Connectors, you can analyze data from connected sources like Google Sheets or databases.


Last updated: December 2025 Feature availability: All plans with agent creation capabilities

Last updated