Two AI agents run my news site; a grounding gate keeps them honest (www.runagentrun.co.uk)

On June 9, 2026, Anthropic unveiled Claude Fable 5, a powerful AI model designed for software engineering and agentic tasks. Within 24 hours, a new news site was born, fully operational with articles researched and published autonomously. This project, led by founder Aaron Coates, showcases a novel workflow involving two AI agents: Claude Fable 5, serving as the supervising architect for quality control, and MiniMax-M3, managing day-to-day content creation. The integration utilizes a git repository for transparency and rollback capabilities, ensuring an accessible audit trail for all editorial activities. This development is significant for the AI/ML community as it demonstrates the potential of using distinct models for different functions, effectively distributing tasks between high-capacity and cost-efficient AI tools. The project emphasizes the importance of establishing validation mechanisms—like stringent fact-checking and schema enforcement—in AI workflows, which help prevent errors such as fabricated information. The innovative approach promises to lower operational costs significantly while improving efficiency, offering a blueprint for other small businesses looking to integrate AI into their content and knowledge workflows. Moreover, the experiment's ongoing nature raises crucial questions about maintaining content quality over time, making it a compelling case study for future AI applications.