Online Resource 1 for "Evidence-monotone authorization: runtime claim control for scientific decision-support software"

Reproducibility supplement for the manuscript. Contains the frozen 28-case OLV authorization benchmark and baseline results, the mechanism-ablation script, the exhaustive chain and non-chain policy-space census with an independent greatest-lower-bound oracle and injected faults, the runtime overhead and memory benchmarks, the state-descriptor dependency audit, the mapping-review template, the 0.6.8/0.6.9 authorization-path digest comparison, and SHA-256 checksums for every file. Version 1.1 adds the cross-domain external-data validation of Section 4.6: frozen EPA AQS and USGS ComCat snapshots with their queries, access times and digests, a second EPA vintage for the observed freshness test, the evidence-to-cap policies hashed before any data was retrieved, a standalone reference implementation that shares no code with the census harness, and reproduce_all.py, which regenerates every count offline and re-checks the policy freeze. Reproduces Tables 3–6 and Figures 2 and 3. See README.txt and REVIEWER_QUICKSTART.md.