gpt-review

A Python based CLI and GitHub Action to use Open AI or Azure Open AI models to review contents of pull requests.

How to install CLI

First, install the package via pip:

pip install gpt-review

GPT API credentials

You will need to provide an OpenAI API key to use this CLI tool. In order of precedence, it will check the following methods:

Presence of a context file at azure.yaml or wherever CONTEXT_FILE points to. See azure.yaml.template for an example.

AZURE_OPENAI_API_URL and AZURE_OPENAI_API_KEY to connect to an Azure OpenAI API:

export AZURE_OPENAI_API=<your azure api url>
export AZURE_OPENAI_API_KEY=<your azure key>

OPENAI_API_KEY for direct use of the OpenAI API
```
export OPENAI_API_KEY=<your openai key>
```
AZURE_KEY_VAULT_URL to use Azure Key Vault. Put secrets for the url at azure-open-ai and the API Key azure-openai-key, then run:
```
export AZURE_KEY_VAULT_URL=https://<keyvault_name>.vault.azure.net/
az login
```

Main Commands

To show help information about available commands and their usage, run:

gpt --help

To display the current version of this CLI tool, run:

gpt --version

Here are the main commands for using this CLI tool:

1. Ask a Question

To submit a question to GPT and receive an answer, use the following format:

gpt ask "What is the capital of France?"

You can customize your request using various options like maximum tokens (--max-tokens), temperature (--temperature), top-p value (--top-p), frequency penalty (--frequency-penalty), presence penalty (--presence-penalty), etc.

Ask a Question about a File

To submit a question to GPT with a file and receive an answer, use the following format:

gpt ask --files WordDocument.docx "Summarize the contents of this document."

2. Review a PR

To review a PR, use the following format:

gpt github review \
    --access-token $GITHUB_ACCESS_TOKEN \
    --pull-request $PULL_REQUEST_NUMBER \
    --repository $REPOSITORY_NAME

3. Generate a git commit message with GPT

To generate a git commit message with GPT after having added the files, use the following format:

git add .

gpt git commit

For more detailed information on each command and its options, run:

gpt COMMAND --help

Replace COMMAND with one of the main commands listed above (e.g., 'ask').

Developer Setup

To install the package in development mode, with additional packages for testing, run the following command:

pip install -e .[test]

Metadata

Release files for gpt-review 0.9.5

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for gpt-review 0.9.5
File	Size	Uploaded	Provenance
gpt_review-0.9.5.tar.gz	34.7 kB	Sep 14, 2023	Without provenance	Details

Built distribution (wheel)

Table of built distributions (wheels) for gpt-review 0.9.5
File	Interpreter	ABI	Platform	Reset
gpt_review-0.9.5-py3-none-any.whl 23.0 kB Sep 14, 2023 Without provenance	Python 3	none	any	Details

Total release size: 57.8 kB

Release files / gpt_review-0.9.5.tar.gz

Download URL	gpt_review-0.9.5.tar.gz
Size	34.7 kB
Tags	Source
SHA-256 checksum How to use checksums	`c9a9e48ec7a561bce38d98ced8a87d54ec831c3553d263ed724607292460ffd5`
BLAKE2b-256 checksum How to use checksums	`67c73759e37e09f7597ea07c4c8fcfbf1ae43e0e43c24fa92911cb06b71bd9c5`
Upload date	Sep 14, 2023
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`python-requests/2.31.0`

Release files / gpt_review-0.9.5-py3-none-any.whl

Download URL	gpt_review-0.9.5-py3-none-any.whl
Size	23.0 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`04a4a996f3742715ffef18ab3f9d847766285d341b9bdd179bc2100fcaf1be73`
BLAKE2b-256 checksum How to use checksums	`5bf36dd9c4ee43adeda507e70b65361d5f6cc5b30a2ce2ee2a53e0fce65e6fbb`
Upload date	Sep 14, 2023
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`python-requests/2.31.0`

Security for gpt-review 0.9.5

Report project as malware

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.