gptdiff 0.1.34

A tool to generate and apply git diffs using LLMs

GPTDiff

Table of Contents

• Quick Start
• Example Usage
• Basic Usage
• Simple Command Line Agent Loops
• Why Choose GPTDiff?
• Core Capabilities
  • CLI Excellence
  • Magic Diff Generation
  • Smart Apply System
• Get Started
  • Installation
  • Configuration
• Command Line Usage

🚀 Create and apply diffs with AI
Modify your project using plain English.

More documentation at gptdiff.255labs.xyz

Quick Start

1. Install GPTDiff

Example Usage of gptdiff

Apply a Patch Directly

gptdiff "Add button animations on press" --apply

✅ Successfully applied patch

Generate a Patch File

gptdiff "Add API documentation" --call

🔧 Patch written to diff.patch

Generate a Prompt File Without Calling LLM

gptdiff "Improve error messages"

📄 LLM not called, written to prompt.txt

---

Basic Usage

cd myproject
gptdiff 'add hover effects to the buttons'

Generates a prompt.txt file containing the full request. Copy and paste its content into your preferred LLM (e.g., ChatGPT) for further experimentation.

Simple command line agent loops

while
do
  gptdiff "Add missing test cases" --apply
done

Requires reasoning model

Why Choose GPTDiff?

• Describe changes in plain English
• AI gets your whole project
• Auto-fixes conflicts
• Keeps code functional
• Fast setup, no fuss
• You approve every change
• Costs are upfront

Core Capabilities

⚡ CLI Excellence

• Target Specific Files - Change just what you need
• Live Updates - See changes as they're made
• Try Before Applying - Test changes safely first
• Clear Pricing - Know costs upfront
• Preview Changes - See what will change with --call
• Fix Mistakes - Automatic error correction

✨ Magic Diff Generation

gptdiff "Convert class components to React hooks" --model deepseek-reasoner

• Full project context awareness
• Cross-file refactoring support
• Automatic conflict prevention

🧠 Smart Apply System

Git-native Workflow:

# 1. Apply AI-generated changes
gptdiff "Improve error handling" --apply

# 2. Review each change interactively
git add -p

# 3. Commit or discard
git commit -m "Enhanced error handling"
git reset --hard  # To undo all changes

gptdiff "Refactor authentication to use OAuth 2.0" --apply

✅ Successfully applied complex changes across 5 files

Get Started

Installation

Requires Python 3.8+. Install from PyPI:

pip install gptdiff
pip install tiktoken  # For token counting

Development install (no pip package yet)

python setup.py install

Configuration

First sign up for an API key at https://nano-gpt.com/api and generate your key. Then configure your environment:

Linux/MacOS

export GPTDIFF_LLM_API_KEY='your-api-key'
# Optional: For switching API providers
export GPTDIFF_MODEL='deepseek-reasoner'  # Set default model for all commands
export GPTDIFF_LLM_BASE_URL='https://nano-gpt.com/api/v1/'

Windows

set GPTDIFF_LLM_API_KEY=your-api-key
rem Optional: For switching API providers 
set GPTDIFF_MODEL=deepseek-reasoner
set GPTDIFF_LLM_BASE_URL=https://nano-gpt.com/api/v1/

The default base URL points to nano-gpt.com's API. Supported models can be specified with:

gptdiff 'your prompt' --model deepseek-reasoner
# Default model can be set via GPTDIFF_MODEL environment variable

OpenAI will not be called unless you specify --call or --apply

Prevent files being appended to the prompt by adding them to .gitignore or .gptignore

Command Line Usage

gptpatch: Apply Diffs Directly

gptpatch is a companion command-line tool to GPTDiff that applies unified diffs directly to your project.

Usage

Apply a diff provided directly:

gptpatch --diff "<diff text>"

Or apply a diff from a file:

gptpatch path/to/diff.patch

Options

• --project-dir: Specify the target project directory (default: current directory)
• --model: (Optional) Specify the LLM model for advanced conflict resolution
• --max_tokens: (Optional) Define the maximum token count for LLM responses during patch application
• --nobeep: Disable the completion beep notification

Workflow

gptpatch first attempts to apply the diff using standard patch logic. If that fails, it automatically falls back to a smart apply mechanism that leverages AI-powered conflict resolution.

For more details, see the gptpatch documentation on our docs site.

After installing the package, use the gptdiff command in your terminal. Change directory into your codebase and run:

gptdiff '<user_prompt>'

any files that are included in .gitignore are ignored when generating prompt.txt.

Specifying Additional Files

You may supply extra files or directories as arguments to the gptdiff command. If omitted, the tool defaults to the current working directory, excluding those matching ignore rules.

Autopatch Changes

You can also call openai and automatically apply the generated git diff with the --apply flag:

gptdiff '<user_prompt>' --apply

Dry-Run Validation

Preview changes without applying them by omitting the --apply flag when using --call:

gptdiff "Modernize database queries" --call

i️ Diff preview generated - review changes before applying

This often generates incorrect diffs that need to be manually merged.

Smart Apply

For robust handling of complex changes, use smartapply. It processes each file’s diff individually via the LLM, ensuring nuanced conflict resolution.

Completion Notification

Use the --nobeep option to disable the default completion beep:

gptdiff '<user_prompt>' --nobeep

Local API Documentation

Preview API docs locally using MkDocs:

pip install .[docs]
mkdocs serve

Open http://localhost:8000 to view the documentation

Python API

Integrate GPTDiff directly into your Python workflows:

from gptdiff import generate_diff, smartapply
import os

os.environ['GPTDIFF_LLM_API_KEY'] = 'your-api-key'

# Create files dictionary
files = {"main.py": "def old_name():\n    print('Need renaming')"}

# Generate transformation diff using an environment string built from the files dictionary
environment = ""
for path, content in files.items():
    environment += f"File: {path}\nContent:\n{content}\n"

diff = generate_diff(
    environment=environment,
    goal='Rename function to new_name()',
    model='deepseek-reasoner'
)

# Apply changes safely using the files dict
updated_files = smartapply(diff, files)

print("Transformed codebase:")
print(updated_files["main.py"])

Batch Processing Example:

from gptdiff import generate_diff, smartapply

files = load_your_codebase()  # Dict of {path: content}

transformations = [
    "Add python type annotations",
    "Convert string formatting to f-strings",
    "Update deprecated API calls"
]

for task in transformations:
    files = smartapply(generate_diff(build_environment(files), task), files)

Integration Note: GPTDiff leverages the AI Agent Toolbox for seamless tool usage across AI models and frameworks, making it ideal for both single responses and complex agent workflows.

Core Functions

• generate_diff(environment: str, goal: str, model: str) -> str
  Generates a git diff implementing the requested changes

  model parameter defaults to GPTDIFF_MODEL environment variable

• smartapply(diff_text: str, environment_str: str, model: str) -> str
  Applies complex diffs while preserving file context

Testing

To run the test suite:

pip install -e .[test]
pytest tests/

This will execute all unit tests verifying core diff generation and application logic.

plangptdiff: Generate plan prompts that call GPTDiff

plangptdiff scans your repo with ripgrep, selects only the files likely to change (always including anything named schema), and writes a ready‑to‑paste prompt to planprompt.txt:

# Prompt only
plangptdiff "add validation to the signup form"

# Prompt that will auto‑apply the diff
plangptdiff "upgrade to Django 5" --apply

The file list is appended to the generated gptdiff command so the LLM sees only the files that matter, keeping prompts lean and costs down.