Raw Capture
Directory pages, coach profile HTML, and team directory snapshots written outside Git.
raw/coaching-tree-app/<RUN_ID>/
Public-facing snapshot of the new coaching-tree.app migration pipeline built on May 6, 2026. This page shows what today's development run captured, normalized, and validated before any full-scale graph import.
The sample run wrote raw snapshots to local storage, extracted structured staging rows, normalized canonical graph files, and produced validation manifests in a reproducible repository layout.
The migration pipeline separates raw crawl artifacts from reproducible repository data so that large HTML snapshots stay local while compact graph-ready files remain versionable.
Directory pages, coach profile HTML, and team directory snapshots written outside Git.
raw/coaching-tree-app/<RUN_ID>/Directory coaches, parsed profiles, career history rows, worked-under rows, team rows, and source metadata.
data/nfl-coaching-tree/staging/Normalized coaches, franchises, aliases, staff tenures, edges, claims, and source-backed graph import inputs.
data/nfl-coaching-tree/canonical/Validation summaries, unresolved lists, pipeline manifests, and optional Neo4j export bundles.
validation/, manifests/, exports/<RUN_ID>/These examples are drawn directly from today's parsed profile sample and show the kind of canonical coach records now produced by the pipeline.
The sample tenure table below shows how coach-season-role rows are separated from narrative profile pages into graph-friendly records.
Current unresolved items come from the intentionally partial crawl window, not from fabricated data. They stay visible until a broader run resolves them.
Today's sample run used the public coaching-tree.app directory and a small coach profile subset.