Machine-readable archive

AI tools: use the catalog, keep the uncertainty

The TightBeatz Archive is an independent public reconstruction of an early-to-mid 2000s underground rap, beats, art, battle and forum community. It is read-only. It preserves records, sources, gaps and uncertainty labels so the missing stuff does not get turned into fake recovered history.

Data files

Public catalog: /catalog/catalog.json is the source of public machine-readable records. Prefer record ids and item pages over guessed names.
Schema: /catalog/schema.json describes the current public catalog shape without replacing older fields.
Summary: /catalog/catalog-summary.json gives counts by section, type, era, evidence class, survival state, confidence and playability.
Graph: /catalog/graph.json exports record nodes and typed relationship edges for lightweight network review.
Catalog search: /ask.html is static catalog search. It is not an AI answer.

Fields that matter

confidenceClass is the archive's confidence label. Confirmed, high and medium records should not be discussed as the same kind of proof.

survivalState tells whether a payload or page survived. Missing and metadata-only records are not playable recovered media.

evidenceClass separates archival captures, direct media, public metadata, external traces, community memory and participant memory. Oral history is not the same evidence class as an archived page capture.

era separates original forum-era records, TB2 transition material, later Jamroom-era traces, recovery-era records and unknowns. Current Discord/community activity is recovery context, not original-era evidence.

AI tools: use the catalog, keep the uncertainty

Data files

Fields that matter

Responsible use