Show HN: TetherDust – Self-hosted AI Analytics Engineer (open source) (github.com)

TetherDust has launched as an open-source self-hosted AI analytics engineer, designed to streamline data interactions between codebases and databases through containerized Model Context Protocol (MCP) servers. The platform excels in creating detailed documentation from both code and database schemas and allows AI agents to generate verifiable SQL queries and dynamic visualizations using d3.js. Notably, TetherDust prioritizes security with strict read-only query boundaries and role-based access controls, ensuring that users operate within their predefined scopes and that all actions are logged immutably. This innovation is significant for the AI/ML community as it facilitates enhanced data accessibility and interaction, empowering teams to leverage AI's capabilities for real-time data analytics without compromising security. TetherDust supports various databases like PostgreSQL and MySQL while being adaptable to any agent that complies with MCP specifications. Its architecture allows seamless integration of AI agents, enabling them to run scheduled queries, generate alerts, and provide contextual responses to natural language queries—all while maintaining stringent oversight on data stability and governance. This positions TetherDust as a versatile solution for organizations looking to harness the potential of AI-driven analytics safely and efficiently.