diff --git a/README.md b/README.md
index de65759..436e7d1 100644
--- a/README.md
+++ b/README.md
@@ -124,6 +124,12 @@ this project sells.
 | Examples | ✅ 5 builds, all 48/48, committed evidence |
 | Comparison gallery | ✅ executed 2026-08-29 — 8 briefs × 2 skills, scored + vision-passed, losses published |
 
+**On the 1.0.0 tag:** it is cut on the measured result — the moat check passed
+decisively (the thesis), while the headline margin (+4.5 vs the ≥5 target) and
+the Hallmark-side S1 (0.39 vs ≥0.50) missed and stand published in the
+gallery. No metric was altered to reach the tag; 1.0.0 is the
+API-stability promise, not a claim that every marketing target was hit.
+
 ## Development
 
 ```bash
diff --git a/package.json b/package.json
index 70a0e2e..8b6007d 100644
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@getpipher/keystone",
-  "version": "0.1.0",
+  "version": "1.0.0",
   "description": "Anti-AI-slop design skill with an executable gate engine. Beats Hallmark by enforcing its gates instead of imagining them.",
   "keywords": [
     "pi-package",
diff --git a/skills/keystone/SKILL.md b/skills/keystone/SKILL.md
index 75d0f9f..f8f5285 100644
--- a/skills/keystone/SKILL.md
+++ b/skills/keystone/SKILL.md
@@ -1,7 +1,7 @@
 ---
 name: keystone
 description: "Anti-AI-slop design skill with an executable gate engine. Use when building a new landing page or web page, auditing an existing site, or when the user invokes Keystone by name. Beats Hallmark by enforcing its gates with a real engine instead of imagining the render."
-version: 0.1.0
+version: 1.0.0
 ---
 
 # Keystone
diff --git a/test/compare/gallery/PROTOCOL.md b/test/compare/gallery/PROTOCOL.md
index c74f50f..6b94758 100644
--- a/test/compare/gallery/PROTOCOL.md
+++ b/test/compare/gallery/PROTOCOL.md
@@ -133,3 +133,8 @@ table with the note that ex-stamp Hallmark was clean there. Vision S1 came in
 BELOW the Hallmark ≥ 0.50 target (0.39) — published as measured. The criterion
 the thesis rests on — the moat check — passed with a wide margin. The overall
 verdict: 3 of 4 criterion-halves met; 1.0.0 NOT cut on this run.
+
+**Update 2026-08-30:** v1.0.0 was subsequently tagged on this measured result —
+the misses above stand as published and no metric was altered to reach the
+tag. The 1.0.0 criterion was API stability (engine, skill, CLI complete and
+green), not the marketing targets.