Skip to content

tool-usage-examples.md

Latest commit

 

History

History
174 lines (123 loc) · 3.85 KB

tool-usage-examples.md

File metadata and controls

174 lines (123 loc) · 3.85 KB

Tool Usage Examples

This document provides detailed examples of how to use each tool in the collection.

CLI Tools

GitHub CLI Demo (cli-demo.gpt)

Purpose: Interact with GitHub repositories using the GitHub CLI

Usage:

gptscript tools/cli/cli-demo.gpt

Example Interactions:

  • List repositories
  • Create issues
  • Manage pull requests
  • View repository information

Data Processing Tools

Local Files Assistant (local-files.gpt)

Purpose: Process and analyze local data files including CSV, JSON, PDF, and Excel files

Usage:

# Process files in sample-data directory
gptscript --workspace=./sample-data tools/data-processing/local-files.gpt

Supported File Types:

  • CSV files: Query, filter, and analyze tabular data
  • Excel files: Read and process spreadsheet data
  • PDF files: Extract and analyze text content
  • JSON files: Parse and query structured data

Example Tasks:

  • "Analyze the sales data in yoy_sales.xlsx"
  • "Extract contact information from key_contacts.txt"
  • "Summarize the content of promotions.pdf"

Workspace Context Tool (context-tool.gpt)

Purpose: Provides workspace context and file system operations

Usage: Typically used as a dependency by other tools

# Used internally by local-files.gpt
Context: ./context-tool.gpt

Social Media Tools

Twitter Workflow (tweets-workflow.gpt)

Purpose: Summarize tweets and extract hyperlinks from Twitter URLs

Usage:

gptscript tools/social/tweets-workflow.gpt

Features:

  • Summarizes tweets in markdown format
  • Extracts and summarizes hyperlinks from tweets
  • Processes tweets synchronously in order
  • Outputs summaries to tweets.md

Current Configuration:

  • Processes specific tweet URLs (can be modified in the tool)
  • Generates structured markdown output with references

Weather Tools

Weather Bot (weather-tool.gpt)

Purpose: Get weather information using WeatherStack API

Prerequisites:

  • WeatherStack API key (set as environment variable)

Usage:

gptscript tools/weather/weather-tool.gpt

Features:

  • Current weather conditions
  • Weather forecasts
  • Location-based weather data
  • Interactive confirmation before API calls

Setup:

  1. Get API key from WeatherStack
  2. Set environment variable: WS_API_KEY=your_api_key
  3. Run the tool

Advanced Usage

Chaining Tools

You can create workflows that combine multiple tools:

# Example: Process data then generate weather report
gptscript --workspace=./sample-data tools/data-processing/local-files.gpt
# Then use the processed data with other tools

Custom Workspaces

Use different data directories:

# Use custom data directory
gptscript --workspace=/path/to/your/data tools/data-processing/local-files.gpt

Makefile Shortcuts

Use the provided Makefile for common operations:

# Install all requirements
make install-requirements

# Run specific demos
make run-cli-demo
make run-api-demo
make run-local-files-demo
make run-tweets-workflow

# List WeatherStack API operations
make ws-list-ops

# Get help
make help

Troubleshooting

Common Issues

  1. Missing Dependencies: Run make install-requirements
  2. API Key Issues: Ensure environment variables are set correctly
  3. File Path Issues: Use absolute paths or ensure correct working directory
  4. Permission Issues: Check file permissions for scripts

Debug Mode

Add verbose output to any tool:

gptscript --verbose tools/cli/cli-demo.gpt

Extending Tools

Adding New Tools

  1. Create tool file in appropriate tools/ subdirectory
  2. Add Makefile target in config/Makefile
  3. Update documentation
  4. Test with sample data

Tool Structure

Each GPTScript tool should include:

  • Clear name and description
  • Required tools and dependencies
  • Parameter definitions
  • Usage instructions