phanindraintelligenzit-afk/ai-workflow-cleanup

AI Workflow Cleanup is a tool built for small and medium businesses that automated too many processes and ended up with broken results. It connects to your existing automation setups built in n8n, Zapier, or Make, scans them for problems, and walks you through fixing what is wrong. The scanner checks each workflow's steps against a list of common failure patterns, things like missing triggers, bad login credentials, timeouts, rate limits, deprecated APIs, and broken data mappings. Every workflow gets a health score between 0 and 100 percent based on how many issues it has and how serious they are. Once problems are found, you can apply a one click automatic fix for known issues, or follow step by step guidance to fix things manually. Every scan and every fix is logged with a timestamp so you have a record of what changed and when. Under the hood, the project is a web application with a Python based FastAPI backend, a React frontend, and a SQLite database accessed asynchronously. The backend exposes a REST API for listing, creating, scanning, and fixing workflows, and the frontend talks to that API to display results. The whole thing can be configured through environment variables, such as the database connection string and allowed origins for cross origin requests. The easiest way to run it is with Docker Compose, which starts both the frontend and backend with a single command, after which you open the app in your browser. It can also be run without Docker by setting up a Python virtual environment for the backend and running the frontend separately with its own package manager. The project currently supports n8n, Zapier, and Make workflows, and its roadmap includes importing workflow files directly, richer AI powered fix suggestions, and exporting fixed workflows back to their original platform. It is released under the MIT license.