P7: differential-testing harness (simurg compare) and nightly run on the Pi

- compare/core: EXACT/NEAR/DISCREPANCY/UNKNOWN with the plan §16.3 tolerances (abs, float32 ULP), NaN = NaN,
  locus classes (assay, variant type, chromosome group, intensity-only); fields absent on both sides are
  not applicable
- compare/records: GTC vs GTC (per-locus arrays, transforms per bin, controls, sample statistics; metadata
  listed, not counted) and VCF vs VCF (records keyed incl. ID; SNV/MNV/indel, merged, multi-allelic classes)
- simurg compare A B (files or folders paired by name): summary.json, discrepancies.tsv.gz (never truncated),
  self-contained dashboard.html; --tolerance, --fail-on-discrepancy (exit 3)
- scripts/nightly.py + install_nightly.sh (crontab 02:30): unit tests, GSA vs gtc2vcf (GTC + VCF), VCF
  regression vs golden, Omni2.5 subset vs vendor GTCs; first run PASS on every step
- tests/unit/test_compare.py; docs/validation/P7.md, README

README.md

@@ -21,9 +21,12 @@ sites, per-sample VCF.gz + tabix or a cohort BCF + CSI, optional INFO/intensity
 auxiliary loci) and `simurg export plink|bedgraph|finalreport|locus-summary` (PLINK 1 `.bed` and PLINK 2 `.pgen`).
 On the GSA demo, genotypes match gtc2vcf and GTCtoVCF with zero unexplained differences and PLINK files match plink2
 (`docs/validation/P6.md`).
+Phase **P7 (differential testing)** is complete: `simurg compare` (EXACT/NEAR/DISCREPANCY/UNKNOWN per field and
+locus class, every discrepancy listed, HTML dashboard) and a nightly run on the Pi (GSA vs gtc2vcf, Omni2.5 subset vs
+vendor GTCs, VCF regression) — `docs/validation/P7.md`.
 Available commands: `simurg version`, `simurg inspect <files…> [--json]`, `simurg ingest`, `simurg normalize`, `simurg genotype`,
 `simurg qc controls|samples|contamination|relatedness|loci|report|suggest-thresholds`, `simurg gtc-to-vcf`,
-`simurg export plink|bedgraph|finalreport|locus-summary`.
+`simurg export plink|bedgraph|finalreport|locus-summary`, `simurg compare`.
 
 ## Quick start