Shippie is an extendable code review and QA agent designed to integrate into CI/CD pipelines and local development workflows. It leverages Large Language Models (LLMs) to identify common code issues like exposed secrets, inefficient code, and potential bugs, aiming to improve shipping velocity and code quality for developers.

How It Works

Shippie functions as a vertically integrated agent within CI/CD pipelines, acting as a human code reviewer. It utilizes LLMs to analyze code, identifying specific types of vulnerabilities and inefficiencies. It also serves as a Model Context Protocol (MCP) client, enabling integration with external tools for broader workflow automation.

Quick Start & Requirements

• Install dependencies using bun i (or npm/pnpm).
• Set up an OpenAI API key in the .env file.
• Run locally with bun start.
• See setup instructions for CI/CD integration.

Highlighted Details

• Reviews code for exposed secrets, inefficiency, and potential bugs.
• Can be run locally on staged files via CLI.
• Written in TypeScript and Bun for a fast CLI experience.
• Acts as an MCP client for external tool integration.

Maintenance & Community

The project uses release-please for automated releases based on Conventional Commits. Discussions for ideas and roadmap items are available in the project's discussion tab.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README.

Limitations & Caveats

The project relies on OpenAI's API, incurring associated costs. The README does not detail specific LLM models used or provide performance benchmarks.