glosa — Rigour Without Infrastructure: A Standalone Scholar Methodology for Human–AI Knowledge Co-Production

v0.1.0 — K0 public working release (v0.2.0: glosa applied to itself — concept paper DOI 10.5281/zenodo.22307841, 48 VERIFIED citation cards, cross-vendor reviewed claim cards; rules 15-17 and gate rule 9 added) — timestamped and citable, NOT peer reviewed, no independent check yet (K1 in glosa requires a cross-vendor I3 check that has not run), tier Dr. Methodology + skill + tools for co-producing rigorous knowledge with AI without institutional infrastructure. Every claim answers five questions bound to the Readout Condition. No priority claim.v0.3.0 (2026-09-05). glosa reviewed its own knowledge base (324 knowledge cards judged against the Readout Universe / Readout Genesis base), ran a literature review on eight improvement questions (60 verified citations), built an improvement DAG, simulated gate precision/recall on a synthetic corpus, and shipped kernel rules 18–28 (injected-infinity/zero taxonomy, gate-type law, forbidden-word rejection, layer-mismatch, verdict class, composite-quote, hidden-AI-fill, inflated-bearing) with FOUNDATION v0.6. Six design nodes remain pending the founder's decision and are marked as such. K0.