site: the agent-platform case-study card carries the re-measured figures

17 runtime schema patches (15 ALTER TABLE, 2 CREATE TABLE), a ~7,000-line spine, and floors
that hold under every counting method: 20,000+ generated lines, 100+ prompts. Re-measured
against the adopter repository on 2026-09-25.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01M1mRRB1K2WdYy82kntMQdx

Author: claude

www/index.html

@@ -92,9 +92,9 @@ <h3>Wizards of Odd</h3>
         <article class="proof-card">
           <p class="proof-tag">Python + TypeScript &middot; production</p>
           <h3>Draagon AI</h3>
-          <p>From 17 runtime <code>ALTER TABLE</code> patches to zero. One ~5,000-line YAML
-          spine generates ~19,600 lines of drift-checked Python and TypeScript &mdash; 44
-          entities and 106 LLM prompts declared once, on the published packages.</p>
+          <p>From 17 runtime schema patches to zero. One ~7,000-line YAML
+          spine generates more than 20,000 lines of drift-checked Python and TypeScript &mdash; 44
+          entities and more than 100 LLM prompts declared once, on the published packages.</p>
           <p class="proof-links">
             <a href="https://metaobjects.com/case-studies/draagon-ai/">Read the case study&nbsp;&rarr;</a>
           </p>