Automating Content with a Self-Critique Loop

As a member of the WebNutch community, I wanted to share my experience with automating content posting. I've built a bot that judges my content before it posts, using a self-critique loop to ensure only high-quality content is published. This setup consists of 57 nodes and 8 daily triggers, and has been a game-changer for my content strategy. The self-critique loop is the key to avoiding the common pitfall of automated accounts looking too perfect and spammy. Instead, my bot generates a draft, which is then reviewed by a second AI agent against quality criteria. If the draft fails to meet the criteria, it loops back and tries again. If it hits the max retries, it skips that slot and notifies me via Telegram. This way, I'm only posting content that I'd be happy with manually. The bot has 8 content slots throughout the day, each with a different goal, such as wildcard, exploit, experiment, or CTA. It pulls recent engagement data, trend signals, and learning patterns from Airtable before each run, making the content contextual and relevant. The bot is also self-healing, so if something breaks, it fixes it and re-runs. I built this setup using n8n, starting with Claude code and manual canvas work. As I added new content types or modified the critique loop, I used Synta to describe the changes in plain English and rebuild the modified sections. The entire workflow is sanitized and available on GitHub for anyone to explore or adapt. If you're interested in learning more about the self-critique loop, I'd be happy to walk you through it. You can find the workflow in the WebNutch marketplace, and I'm excited to hear your thoughts and feedback from the community.