This is the most useful n8n workflow I have ever built...

Key Concepts

• N8N: A workflow automation tool used to orchestrate the interaction between Slack, GitHub, and local scripts.
• Claude Code (Headless Mode): An AI coding agent that executes code changes, commits, and pushes to repositories without a user interface.
• GitHub Webhooks: Used to trigger N8N workflows automatically when specific events (like issue creation) occur.
• Ngrok: A tunneling service used to expose a local N8N instance to the internet so GitHub can send webhook events to it.
• Structured Output: A technique using OpenAI models to parse natural language (Slack messages) into specific JSON schemas (Issue Title/Body).
• Self-Hosted N8N: Required for executing local system commands (e.g., running Python scripts and Claude Code).

---

1. System Architecture and Workflow

The project consists of two primary N8N workflows that function as a closed-loop autonomous coding assistant:

• Workflow 1 (The Executor): Triggered by a GitHub issue. It executes a local Python script that clones the repository, runs Claude Code in headless mode to implement the requested changes, commits the code, pushes to a new branch, and creates a Pull Request (PR) via the GitHub CLI.
• Workflow 2 (The Interface): Triggered by a Slack message. It uses an OpenAI agent to convert the message into a structured GitHub issue (Title/Body), creates the issue on GitHub, and notifies the user in Slack.

2. Step-by-Step Implementation

A. Environment Setup

1. N8N Installation: Install the self-hosted version (e.g., via npm install -g n8n).
2. Tunneling: Use ngrok http 5678 to create a public URL for the local N8N instance.
3. Environment Variables: Launch N8N with specific flags:
   • WEBHOOK_URL: The Ngrok URL.
   • N8N_ENABLE_UNSAFE_CORE_NODES=true: Required to execute local terminal commands.
   • N8N_RUNNERS_ENABLED=true.

B. The Python Automation Script (handle_issue.py)

The script acts as the bridge between N8N and the local file system:

1. Inputs: Accepts repository name, issue number, title, and body as command-line arguments.
2. Git Operations: Clones the repo (if not present), checks out the main branch, pulls updates, and creates a new feature branch (claude-issue-[number]).
3. AI Execution: Runs claude -p "fix issue #[number] [title] [body] --permission-mode accept-edits".
4. Finalization: Adds changes, commits with the issue number, pushes to the remote branch, and uses gh pr create to open a pull request.

C. Slack-to-GitHub Integration

1. Slack API: Create a Slack App with channels:history, channels:read, and chat:write scopes.
2. Event Subscription: Configure the Slack App to send events to the N8N webhook URL.
3. AI Agent: Use an OpenAI node with a JSON schema to extract issue_title and issue_body from the Slack text.
4. GitHub Action: Use the "Create Issue" operation to post the structured data to the repository.

3. Key Arguments and Perspectives

• Autonomous vs. Assisted: The creator emphasizes that this is not just a prompt-response tool but a "fully autonomous" agent that handles the entire lifecycle of a feature request—from communication to deployment-ready PR.
• Non-Technical Collaboration: A primary use case is enabling non-technical team members (e.g., co-founders) to request UI changes or features via Slack, which the developer then only needs to review and merge.
• Security/Transparency: The use of "unsafe" nodes is necessary for local execution, requiring the user to be aware of the security implications of running automated scripts on their machine.

4. Notable Quotes

• "I'm talking about a workflow where I can talk to an AI agent via Slack or I can just create GitHub issues and then this bot takes the GitHub issue, does the changes, commits them and pushes them to GitHub and creates a pull request."
• "The biggest issue is setting up everything so it works... wiring everything up is not really the biggest issue."

5. Synthesis and Conclusion

This project demonstrates a powerful integration of LLM-based coding agents (Claude Code) with low-code automation (N8N). By bridging Slack and GitHub through a local Python-based execution layer, the system effectively removes the friction of manual coding for minor tasks. The key takeaway is the modularity: by using GitHub issues as the "source of truth" for tasks, the system remains robust, allowing for asynchronous development where the AI handles the heavy lifting of implementation and version control.