treemd

treemd is a command-line tool that assembles multiple files from your project into a single context, primarily designed for use with Large Language Models (LLMs). It generates a markdown representation of your directory structure and file contents, making it easy to provide comprehensive project context to AI models.

Features

• Generate an ASCII tree representation of directory structure
• Includes file contents in the output
• Respects .gitignore rules
• Filters files by extension
• Estimates token count for LLM context sizing

Installation

To install treemd globally, run:

npm install -g treemd

Or use npx to run it without installing:

npx treemd

Usage

Basic usage:

treemd [directory] [options]

If no directory is specified, treemd will use the current directory.

Options

• -e, --extensions <extensions>: Comma-separated list of file extensions to include
• -s, --silent: Suppress token count output

Examples

1. Generate markdown for the current directory:

   treemd

2. Generate markdown for a specific directory, including only JavaScript and TypeScript files:

   treemd /path/to/your/project -e js,ts

3. Copy the output directly to clipboard (macOS) for pasting into an LLM interface:

   treemd | pbcopy

Output

The output is in Markdown format and includes:

1. An ASCII tree representation of the directory structure
2. File contents for text files

Example output:

# File tree
```
project/
├── src/
│   ├── index.js
│   └── utils.js
├── tests/
│   └── test_utils.js
└── README.md
```

# Files content
**index.js:**
```
console.log("Hello, world!");
```

**utils.js:**
```
function add(a, b) {
  return a + b;
}
```

**test_utils.js:**
```
assert.equal(add(2, 2), 4);
```

**README.md:**
```
# My Project
This is a sample project.
```

Caveats

• treemd respects .gitignore files in the scanned directory.
• It always excludes certain files/directories like .git, .gitignore, etc.
• Only text files are included in the output.
• The tool estimates token count, which may vary slightly from the actual count used by specific LLM implementations.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License.