foldup

A command-line tool that "folds" your codebase into a single Markdown file that can be easily passed to LLMs.

The output codebase.md file includes:

1. Your project tree
2. The filepath and contents of each file in the codebase

Here's an example of what the output looks like for a very simple project:

# PROJECT TREE

myproject
├─ .gitignore
├─ README.md
└─ src
   ├─ main.py
   └─ utils.py

# .gitignore

```
node_modules
*.log
```

# README.md

```md
# My Project
Lorem ipsum
```

# src/main.py

```python
def main():
    print("Hello, World!")
```

# src/utils.py

```python
def multiply(a, b):
    return a * b
```

Installation

1. Clone the repository:

git clone https://github.com/nathanclairmonte/foldup.git
cd foldup

2. Install as a CLI tool:

pip install .

Now you should be able to run foldup from anywhere in your system! Please let me know if anything doesn't work though.

Usage

# Process current directory
foldup .

# Process specific directory
foldup /path/to/project

Configuration

You can configure Foldup with command-line flags, a config file, or a .foldignore file.

Command-line Flags

Config File (default: foldup.yaml)

Create a foldup.yaml config file in your project root to customize the behavior of Foldup. This example config file shows all the default settings. You can specify a custom config file with the -c flag.

# Default config
exclude:
  # common files to exclude
  - __pycache__
  - node_modules
  - .git
  - venv
  - .venv
  - .env
  - .env.*
  - .idea
  - .vscode
  - dist
  - build
  - .next
  - coverage

  # foldup-related
  - codebase.md
max_file_size_mb: 2.0
show_processed_files: false
estimate_tokens: false
tree_only: false

.foldignore File

You can also configure foldup to ignore certain directories or files by creating a .foldignore file in your project root. The syntax is the same as .gitignore.

# Example .foldignore
*.log
temp/
*.pyc
.DS_Store

Estimated Token Count

The estimated token count can be optionally displayed by passing the --estimate-tokens flag. Foldup uses the tiktoken library to estimate token count. Remember that this is just an estimate, the actual token count may vary (but probably not by an insane amount).

N.B. Tiktoken uses the GPT-4 tokenizer. For ChatGPT, it should be relatively close. For Claude, it could be off by ±20%.

Development

These steps require that you have the UV project manager installed.

1. Clone the repository:

git clone https://github.com/nathanclairmonte/foldup.git
cd foldup

2. Install dependencies:

uv sync

3. Make any changes you want to the code!

4. Run the CLI tool locally:

uv run python -m src.foldup.cli