EvoOntology is a self-evolving ontology layer that lets data agents actively query and update a schema/content/tool-layer ontology exposed as an MCP server, rather than relying on static, manually built semantic layers. A builder agent constructs the ontology automatically and a self-evolution loop applies attribution-guided typed edits that are accepted only after a backbone-conditional paired evaluation. Across three established data-agent benchmarks and four LLM backbones, EvoOntology consistently outperforms strong baselines and existing semantic-layer approaches.
