Target-level LLM assessments attached to a dictionary

The companion accessor to semantic_suggestions(), for the semantic_llm_assessments attribute that suggest_semantics(llm_assess = TRUE) attaches. Reading LLM review is never itself an LLM call: this only reports assessments that already exist.

Usage

semantic_llm_assessments(x)

Arguments

x

A dictionary carrying the attribute, a list in the shape returned by infer_salmon_datapackage_artifacts() / create_sdp()-style inference, or a length-one character path to a written Salmon Data Package.

Value

A tibble of target-level assessment rows, or NULL when none are attached. A package path always returns NULL — assessments are not written into the package, so a package on disk cannot carry them.

Examples

dict <- tibble::tibble(column_name = "spawner_count")
semantic_llm_assessments(dict)
#> NULL