Planning With Files

What it does

This skill enables AI agents to plan and execute tasks involving files. It allows for sequential steps, each potentially interacting with a file (reading, writing, modifying). The agent can reason about file contents and use this information in its planning process, ultimately leading to complex workflows based on file manipulation.

When to use it

• Automated report generation: Create reports by extracting data from multiple CSV files and combining them into a final document.
• Code refactoring: Plan steps for automated code modifications across several source code files.
• Data processing pipelines: Design workflows that read, transform, and write data between different file formats (e.g., converting JSON to Parquet).
• Document summarization: Extract key information from a series of documents and synthesize it into a concise summary.

Key capabilities

• File reading and writing
• Sequential task planning
• Reasoning about file contents
• File format manipulation (implied)

Example prompts

• "Create a plan to combine the data from sales_data_2023.csv and sales_data_2024.csv into a single CSV file named combined_sales.csv."
• “I have several text files containing customer reviews. Plan how I can extract common themes and write them to a summary document.”
• "Read the contents of config.json, then create a plan to update values in settings.txt based on those configurations."

Tips & gotchas

• Ensure the AI agent has appropriate file access permissions before using this skill.
• Complex planning involving many files can be computationally intensive and may require significant processing time.