Automated AI Code Review and Labeling for GitHub Pull Requests - n8n Workflow

Engineering Managers seeking to enforce consistent coding standards.
DevOps Specialists automating development pipelines.
Developers looking to receive immediate, structured feedback on PRs.
Users who want advanced examples of LangChain Agents within an n8n workflow.

• Anyone interested in powerful n8n templates for software development.

Manual code review is often a bottleneck in the software development lifecycle. This comprehensive n8n workflow solves this by instantly analyzing incoming Pull Request (PR) diffs using an AI Agent powered by GPT-4o-mini. The automation is initiated by a GitHub n8n trigger, minimizing latency.

Crucially, this n8n template utilizes the LangChain Agent framework, which allows the AI to reference external knowledge—specifically, a Google Sheet containing your team's custom coding best practices. This ensures the review is not generic, but tailored and opinionated. After the review is generated, the n8n workflow posts the detailed feedback as a comment and automatically labels the PR, establishing a clear, auditable trail. This advanced n8n solution dramatically speeds up the feedback loop and maintains code quality.

This efficient n8n workflow executes the following steps:

1. Trigger on PR: The n8n workflow begins when the GitHub n8n trigger detects a new pull_request event in the configured repository.


2. Fetch Diffs: An HTTP Request n8n node uses the GitHub API to fetch the specific file changes (diffs) associated with the newly opened PR.


3. Construct Prompt: A dedicated n8n code node processes the raw diff data. It sanitizes the code (handling backticks) and structures the data into a precise natural language prompt, instructing the AI agent on its role (e.g., 'senior iOS developer') and mission.


4. AI Agent Review: The data is passed to the Code Review Agent, an n8n node powered by LangChain. This agent leverages the OpenAI Chat Model (gpt-4o-mini) and utilizes the Google Sheets Tool (Code Best Practices) to perform a contextual review.


5. Post Comment: The resulting AI-generated review is extracted from the agent's output. A GitHub n8n node then posts this detailed review as a comment directly on the originating pull request.


6. Label PR: Finally, the n8n workflow uses another GitHub n8n node to add a status label (e.g., ReviewedByAI) to the PR, marking it as processed by the automated system. This ensures the full cycle is handled entirely within this single n8n template.

To deploy this powerful n8n workflow, follow these steps:

1. Import: Copy the provided JSON and import it into your n8n instance via the 'Workflows' menu.


2. GitHub Trigger Setup: Configure the 'PR Trigger' n8n node. You will need to select your repository and connect your GitHub OAuth2 credentials. Ensure the webhook is correctly set up in your GitHub repository settings to send pull_request events to this n8n trigger URL.


3. Credential Setup: Ensure you have credentials set up for:

1. Activate: Once all connections and parameters are configured, activate the n8n workflow to start monitoring for new pull requests.

This n8n workflow relies on several specialized n8n nodes:

PR Trigger (GitHub Trigger): The starting n8n trigger. It listens for pull_request events on a defined GitHub repository, initiating the entire n8n workflow instantly upon creation or update of a PR.
Get file's Diffs from PR (HTTP Request): Uses the dynamic payload from the n8n trigger to construct an API call to GitHub, retrieving the raw file patches necessary for the review.
Create target Prompt from PR Diffs (Code): A critical n8n code node that transforms raw JSON diffs into a single, highly detailed text prompt, ready for the AI agent to consume. It ensures proper formatting and context.
Code Best Practices (Google Sheets Tool): Functions as a retrieval tool for the AI Agent. The agent can query this n8n node to retrieve specific coding standards defined in the sheet, grounding the review in team policy.
OpenAI Chat Model (LangChain LM ChatOpenAi): Specifies the language model (gpt-4o-mini) used for generating the review content. This is the intelligence backbone of the n8n template.
Code Review Agent (LangChain Agent): This intelligent n8n node orchestrates the entire AI process, deciding when to use the Google Sheet tool and passing the contextual prompt to the language model to generate the final output.
GitHub Robot (GitHub): This n8n node posts the generated review comment back to the original pull request, closing the loop of the automated n8n workflow.
Add Label to PR (GitHub): A final administrative n8n node that ensures the PR is labeled, showing the automated review process has completed.