Show HN: Define MCP tools as YAML specs (github.com)

DeclarAgent has launched a declarative runbook executor for AI agents, allowing for the execution of multi-step YAML workflows in a structured and safe manner. This tool is significant for the AI/ML community as it addresses the inherent risks associated with LLM agents executing commands directly. By enabling YAML plans that are human-readable and version-controllable, DeclarAgent enhances the safety and auditability of AI-driven workflows. Features include dry-run capabilities to preview actions, destructive step gating for controlled execution, and structured JSON results for clarity on operations. The tool features various step types, such as shell commands and HTTP requests, facilitating versatile and complex workflows. Users can validate plans, simulate execution, and run workflows with the added assurance of typed error reporting. Once set up, DeclarAgent automatically exposes YAML files as callable tools over the Model Context Protocol (MCP), allowing for easy integration into existing systems. This innovation not only improves the usability of LLM agents but also encourages their safe deployment in production environments, significantly reducing risk while enhancing functionality.