f2llm 0.1.6

A CLI tool to parse files and create LLM prompts

F2LLM - Folder to LLM CLI Tool

A simple command-line tool for converting folder structures to JSON format for LLM processing and applying LLM-generated changes back to your project.

Installation

Requires Python 3.6+ and the pathspec library:

pip install pathspec

Usage

F2LLM has two main modes:

1. Parse Mode (Default)

Convert a folder structure into JSON format:

python f2llm.py <folder_path> <output.json>

Example:

python f2llm.py my_project project_files.json

This creates a JSON file containing all files in the folder with their content wrapped in code fences.

2. Apply Mode

Apply LLM-generated changes from JSON back to your project:

python f2llm.py --apply <changes.json> <target_folder>

Example:

python f2llm.py --apply llm_response.json my_project

JSON Schema

The tool expects/produces JSON files with this structure:

{
  "modified_files": [
    {
      "file_path": "src/main.py",
      "notes": "Updated function logic",
      "content": "```python\n# file content here\n```"
    }
  ],
  "added_files": [
    {
      "file_path": "src/new_file.py", 
      "notes": "New utility module",
      "content": "```python\n# new file content\n```"
    }
  ],
  "moved_files": [
    {
      "old_path": "old/location.py",
      "new_path": "new/location.py",
      "notes": "Reorganized project structure"
    }
  ],
  "unchanged_files": [
    "README.md",
    "package.json"
  ],
  "deleted_files": [
    {
      "file_path": "obsolete_file.py",
      "notes": "No longer needed"
    }
  ]
}

Features

• Gitignore Support: Automatically respects .gitignore files during parsing
• File Operations: Handles create, modify, move, and delete operations
• Code Fences: Automatically wraps file content with appropriate language syntax highlighting
• Safe Paths: Creates necessary directories when applying changes
• Error Handling: Graceful handling of file read/write errors

Common Workflow

1. Extract your project for LLM analysis:

   python f2llm.py my_project prompt.json
   

2. Send prompt.json to your LLM with instructions

3. Apply the LLM's response back to your project:

   python f2llm.py --apply llm_response.json my_project
   

Notes

• The tool preserves file permissions and creates directories as needed
• Binary files or files that can't be read as UTF-8 will show an error message in their content
• When moving files, the tool creates necessary parent directories automatically