Add rendered evidence for Portland Home Security override

experiments/2026-09-14/portland-home-security/rendered-after-evidence.md

@@ -0,0 +1,58 @@
+# Rendered After-State Evidence — Portland Home Security
+
+## Experiment
+Human/business-rule override of the ĀML prototype gate for the homepage sales-status ticker.
+
+## Evidence source
+User-supplied mobile screenshot captured after the Base44 version was published on 2026-09-14 at approximately 08:39 Pacific Time.
+
+## Screenshot integrity
+- Uploaded artifact: `IMG_3781.jpeg`
+- Bytes: `245288`
+- SHA-256: `bb83fbb5e114042810c9121af8dc6783b528894be07a03331eb3caa32fa14f56`
+
+The binary screenshot is not asserted to be stored in this repository by this record. This document preserves its digest, provenance, and observed rendered state.
+
+## Rendered observations
+The screenshot visibly shows the retained blue sales-status ticker at the top of the mobile homepage with:
+
+- `SALES OPEN NOW`
+- `7 Days a Week · 7:00 AM–10:00 PM Pacific Time`
+- `CALL OR TEXT (971) 458-1814`
+- a separate local-time strip showing Portland, Oregon and Pacific Time
+
+The screenshot also shows the Security First Alarm navigation/contact header, the residential-security hero imagery, and a lower persistent `CALL OR TEXT` action. This confirms that the element selected for the experiment remained rendered after publication.
+
+## Decision chain
+Declared prototype inputs:
+
+```text
+attention_cost = 4
+restoration_value = 2
+```
+
+Prototype gate:
+
+```text
+render_allowed = restoration_value >= attention_cost
+2 >= 4 -> false
+GATE RESULT = SUPPRESS
+```
+
+Final governed disposition:
+
+```text
+RETAIN-BY-BUSINESS-RULE
+```
+
+Reason: the ticker carries distinct operational information—live sales availability and hours—in addition to contact routing. The business-rule review therefore overrides the simple prototype gate.
+
+## Base44 evidence
+- Before checkpoint: `6aa813b67e289e6e7f560d86`
+- Before source commit: `7f890be3ac6402f651d79c0e42d5f635039fa3a3`
+- After checkpoint: `6aa813d7c1e184d9c68e2a2d`
+- After source commit: `e85e2d2d1dfc51e4ab49f687acebdc210f868159`
+- Production build: PASS
+
+## Evidence boundary
+This is project-authored field evidence. The attention/restoration values are declared/modelled policy inputs, not objective cognitive, psychological, neurological, wellbeing, or conversion measurements. The screenshot demonstrates the rendered after-state; it does not by itself establish that retaining the ticker improved user outcomes or commercial performance.