Irontology is a Rust-based enterprise semantic runtime designed for knowledge ingestion, latent ontology discovery, and impact analysis. It processes diverse information sources, including program code, architecture documents, project plans, meeting notes, presentations, diagrams, database schemas, and content from various silos. The system converts these artifacts into evidence-bearing semantic objects, correlates them to uncover shared or conflicting meanings, and exposes the resulting knowledge through MCP for agents and humans to explore impacts and dependencies safely. It aims to surface hidden dependencies, disambiguate terms, and support architectural and operational decision-making by creating a living, evidence-backed semantic model of an organization's scattered knowledge.
主な機能
01Ingests heterogeneous artifacts from various source systems and silos
02Converts artifacts into evidence-bearing semantic objects with provenance
03Correlates semantic objects across silos to discover shared or conflicting meaning
04Exposes queryable knowledge through an MCP decision support surface
05Integrates Rhai for configurable semantic behavior and Python for complex extractors and LLM workflows
060 GitHub stars
ユースケース
01Discover hidden dependencies and relationships across diverse enterprise data sources
02Surface missing or contradictory assumptions within an organization's knowledge base
03Understand the downstream impacts of changes to systems, policies, or data
