dumpcat 0.1.1

Dump a directory's file tree and contents into a single formatted output — built for LLM prompts.

dumpcat

Dump a directory's file tree and contents into a single formatted output — built for LLM prompts.

Highlights

• One command to dump your entire codebase (or a filtered slice of it) into a single output
• Smart defaults — respects .gitignore, skips binaries, limits file sizes
• Three output formats — Markdown, plain text, or JSON
• Filterable — by extension, glob pattern, depth, and file size
• Clipboard ready — copy output directly with -c
• Prompt-friendly — prepend/append text or template files for LLM context
• Configurable — project-level .dumpcat.toml with named profiles
• Zero bloat — one runtime dependency (pathspec), everything else is stdlib

Installation

Requires Python 3.11+.

uv tool install dumpcat

# Or with pip
pip install dumpcat

# Or with pipx
pipx install dumpcat

Quick start

# Dump current directory as Markdown
dumpcat

# Only Python files, 2 levels deep
dumpcat -d 2 -i .py

# Prepend a prompt and copy to clipboard
dumpcat -i .py -c -p "Review this code for security issues"

# Plain text with stats
dumpcat -f plain -s

# Just the tree structure
dumpcat --tree-only

# JSON output for tooling
dumpcat -f json --no-tree

Output

dumpcat renders file contents followed by a file tree:

# File Contents

## `src/main.py`

```python
import sys

def main():
    print("hello")
```

# File Tree

```
.
├── src/
│   ├── main.py
│   └── utils/
│       └── helpers.py
└── README.md
```

See the output formats documentation for Markdown, plain text, and JSON examples.

CLI reference

dumpcat [OPTIONS] [PATH]

Flag	Short	Description
--output PATH	-o	Write output to a file
--clipboard	-c	Copy output to clipboard
--depth INT	-d	Max directory depth
--include EXT	-i	Include only these extensions (repeatable)
--exclude PATTERN	-e	Exclude glob patterns (repeatable)
--gitignore / --no-gitignore		Respect .gitignore rules (default: on)
--tree-only		Show only the tree, no file contents
--no-tree		Show only file contents, no tree
--prepend TEXT	-p	Text or @filepath to prepend
--append TEXT	-a	Text or @filepath to append
--max-size SIZE		Skip files larger than this (default: 1mb)
--stats	-s	Show file count, lines, and estimated tokens
--format FORMAT	-f	markdown, plain, or json (default: markdown)
--config PATH		Path to config file
--profile NAME		Named profile from config
--follow-symlinks		Follow symbolic links
--hidden		Include dotfiles and dotdirs
--line-numbers	-n	Add line numbers to file contents

See the full CLI reference for details.

Configuration

Create a .dumpcat.toml in your project root:

[default]
exclude = ["__pycache__", "*.pyc", ".venv"]

[profiles.python]
include = [".py", ".pyi", ".toml", ".md"]
exclude = ["__pycache__", "*.pyc", ".venv", ".mypy_cache"]

[profiles.web]
include = [".js", ".ts", ".tsx", ".css", ".html"]
exclude = ["node_modules", "dist", ".next"]

dumpcat --profile python

See the configuration documentation for details.

Documentation

Full documentation is available at allenfp.github.io/dumpcat.

License

MIT