SWE Kit: PR Review Agent

Build a PR Review agent with SWE Kit

Overview

SWE Development Kit (swekit) is a powerful framework for building Software Engineering agents using Composio’s tooling ecosystem. This example demonstrates how to build a Pull Request (PR) Review Agent using SWE Development Kit (swekit). The PR Review Agent uses Composio’s tooling ecosystem to help automate and enhance the PR review process.

Getting Started

1. Installation

   Begin by installing the core packages:

   Installing SweKit

   pip install swekit composio-core
   

   For additional functionality, install packages for your preferred framework (e.g., LangChain):

   Installing LangChain and Composio LangChain Plugin

   pip install langchain composio-langchain
   
2. Connect your Github Account

   To utilize Github Issues as a task source, link your Github account as follows:

   Composio CLI

   Use Composio CLI

   composio add github
   

   Locally using Github Auth Token

   Use Github Auth Token

   export GITHUB_ACCESS_TOKEN=<your_token>
   

   There are two ways to provide the GitHub access token for git clone:

   1. Set the environment variable GITHUB_ACCESS_TOKEN='<git_access_token>'.
   2. Use the GitHub account connected to your toolset entity. However, this method has limitations:
      • The agent won’t be able to push or create PRs.
      • You need to set export ALLOW_CLONE_WITHOUT_REPO='true'.

   Option 1 is recommended for full functionality.

3. Create a New Agent

   Generate your agent’s scaffolding:

   Create a new Folder with basic PR Review agent code

   swekit scaffold pr_review -f langgraph -o pr_review_agent
   

   Argument	Description	Accepted Values
   -f	Specifies the framework to use	langgraph
   -o	Sets the output directory for generated files	Any valid directory path

   Scaffold support for other frameworks coming soon!

   This process will establish a new agent at pr_review_agent with essential files:

   • main.py: The main script to execute the agent
   • agent.py: The agent’s core definition
   • prompts.py: Prompts to guide the agent’s actions
   • input.py: A helper file to take inputs from the user
   • tools.py: A file to add custom tools to the agent
4. Start Docker Server

   To use Docker as the default workspace environment, ensure your Docker server is running

   If you prefer to run the agent locally without Docker (Unsafe), modify the workspace configuration in agent.py by setting the WorkspaceType.Docker() to WorkspaceType.Host().

5. Run the Agent

   To activate the agent, proceed to its directory and execute:

   Run the agent

   cd pr_review_agent/agent
   python main.py
   

   You will be prompted to specify the repository and issue for the agent to address.