Corpus & data pipeline
Corpus
A corpus is the full body of data a system works over. In GraphCanon it is every tool, repo, and piece of metadata currently in the graph.
A corpus is the complete collection of documents or records a system is built on. When a signal is described as computed "across the corpus", it means it was derived from the whole set - not a single example.
The word shows up everywhere in AI tooling: models are trained on a corpus, search runs over a corpus, and pipelines keep a corpus fresh.
In GraphCanon
GraphCanon's corpus is every published tool and its GitHub-backed metadata. Batch jobs (ingest, enrich, trust scans) operate over the corpus, and "/llms-full.txt" publishes it in one machine-readable file.
See also
Related terms
Last reviewed 2026-07-09