From 42e4e29ac1c2547bb3a66f826fd9fd5b9fecf2b3 Mon Sep 17 00:00:00 2001
From: 732642856 <732642856@qq.com>
Date: Sat, 18 Jul 2026 01:02:02 +0800
Subject: [PATCH] docs: record jyotishyamitra observation comparison

---
 .../jyotishyamitra_pinned_adapter_probe_2026_07_18.md  | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/docs/research/jyotishyamitra_pinned_adapter_probe_2026_07_18.md b/docs/research/jyotishyamitra_pinned_adapter_probe_2026_07_18.md
index 3cdea47c..ec9aa2e5 100644
--- a/docs/research/jyotishyamitra_pinned_adapter_probe_2026_07_18.md
+++ b/docs/research/jyotishyamitra_pinned_adapter_probe_2026_07_18.md
@@ -57,6 +57,10 @@ Replay stability:
 
 ## Next
 
-1. Extract D1/D2/D4/D9/D10, Shadbala/Balas, AshtakaVarga, and Vimshottari fields.
-2. Compare field-by-field against local, Xalen, PyJHora, and jyotishganit as independent observation only.
-3. Keep `promotion_allowed=false` unless external worked examples arbitrate method variants.
+The D1/D2/D4/D9/D10 sign surface now has 45 observation rows against archived local and Xalen raw:
+
+- local: `38/45` matches;
+- Xalen: `28/45` matches; Xalen's absent Rahu/Ketu rows are `not_comparable`, not mismatches;
+- D2 is a method-sensitive difference cluster and remains unarbitrated.
+
+Balas, AshtakaVarga, and Vimshottari remain raw-only until their units, schema, and method contracts are independently aligned. Keep `promotion_allowed=false` unless external worked examples arbitrate method variants.