# Metadata Grounding Metadata grounding uses attributes such as source, date, author, collection, ownership, scope, lineage, permissions, and document type to guide retrieval and interpretation. AutoSites' account folders, frontmatter, source links, editorial status, and collection routes are not incidental metadata; they are part of the meaning of the content. In an AI system they can constrain search, distinguish canonical from derived records, and preserve provenance. Related: [[wiki/Context Engineering|Context Engineering]], [[wiki/Provenance|Provenance]], [[wiki/Layered Context Architecture|Layered Context Architecture]]. Source: [OpenAI — Inside our in-house data agent](https://openai.com/index/inside-our-in-house-data-agent/)