Infer the machine-readable MCP coverage state for a document payload.
The requestedDate should be the current analysis/run date (i.e. "today"), NOT the document's own publication date. Only when the document was published on the same day as the analysis run and still lacks full text do we infer not_indexed (indexing lag). For older documents without full text, the correct state is metadata_only.
Infer the machine-readable MCP coverage state for a document payload.
The
requestedDateshould be the current analysis/run date (i.e. "today"), NOT the document's own publication date. Only when the document was published on the same day as the analysis run and still lacks full text do we infernot_indexed(indexing lag). For older documents without full text, the correct state ismetadata_only.