diff --git a/docs/prompt-pool-routing-map.md b/docs/prompt-pool-routing-map.md
index 860b3fd..30cde8a 100644
--- a/docs/prompt-pool-routing-map.md
+++ b/docs/prompt-pool-routing-map.md
@@ -727,6 +727,10 @@ Important POV rule:
 
 - In POV rows, location anchors must stay behind, beside, or at the frame edges.
   The foreground belongs to the POV body/hands and visible partner/action.
+- Profile `foreground` anchors such as desk edges, counters, shelves, doors, or
+  pillars are normal non-POV placement aids. POV camera text should not reuse
+  them as viewer-side objects; it should reserve the lower foreground for POV
+  body or hand cues.
 
 ## Formatter Routes
 
diff --git a/scene_camera_adapters.py b/scene_camera_adapters.py
index da557a5..c43ab96 100644
--- a/scene_camera_adapters.py
+++ b/scene_camera_adapters.py
@@ -937,10 +937,10 @@ def scene_distance_detail(
     subject, _pronoun = scene_subject_terms(subject_kind, pov_labels)
     if pov_labels:
         if "wide" in distance or "full-body" in distance or "full body" in distance:
-            return f"wide POV keeps {subject} readable with {profile['place']} context behind them"
+            return f"wide POV keeps {subject} readable with {profile['place']} context behind them and environmental anchors only beside or beyond the action"
         if "close" in distance:
             return f"close POV keeps {subject} dominant with {profile['place']} context only at the sides or background"
-        return f"medium POV keeps {subject} dominant with room context behind them"
+        return f"medium POV keeps {subject} dominant with room context beside or behind them"
     if "wide" in distance or "full-body" in distance or "full body" in distance:
         return f"wide crop keeps the {profile['foreground']}, {profile['midground']}, and {profile['background']} readable"
     if "close" in distance:
@@ -969,7 +969,7 @@ def scene_elevation_detail(
         if "low-angle" in elevation:
             return f"low angle keeps POV body cues low while the {profile['background']} rises behind {pronoun}"
         if "elevated" in elevation:
-            return f"elevated POV keeps the viewer's eye line slightly higher than {subject}, with {profile['foreground']} only behind or at the side edges"
+            return f"elevated POV keeps the viewer's eye line slightly higher than {subject}, with location anchors only beside or behind {pronoun}"
         if "high-angle" in elevation:
             return f"high angle looks down from the viewer's position with {profile['midground']} only in the background"
         return f"eye-level angle keeps {profile['midground']} behind {pronoun}"
@@ -1022,7 +1022,8 @@ def scene_camera_directive(
     if pov_labels:
         return (
             f"{profile['layout_label']} from POV{geometry_clause}: {direction_detail}. "
-            f"{distance_detail}; {elevation_detail}; use the multiangle camera only as first-person spatial geometry."
+            f"{distance_detail}; {elevation_detail}; lower foreground is reserved for POV body or hand cues; "
+            f"use the multiangle camera only as first-person spatial geometry."
         )
     return (
         f"{profile['layout_label']}{geometry_clause}: {direction_detail}; "
diff --git a/tools/prompt_smoke.py b/tools/prompt_smoke.py
index 898f8c4..2cafdc2 100644
--- a/tools/prompt_smoke.py
+++ b/tools/prompt_smoke.py
@@ -6018,6 +6018,8 @@ def smoke_insta_pair_camera_split() -> None:
     _expect("back-right quarter view" in hard_scene, "hard camera scene missed hard orbit direction")
     _expect("low-angle shot" in soft_scene, "soft camera scene missed soft elevation")
     _expect("elevated shot" in hard_scene, "hard camera scene missed hard elevation")
+    _expect("near desk edge" in soft_scene, "non-POV camera scene lost coworking foreground anchor")
+    _expect("laptop corner" in soft_scene, "non-POV camera scene lost coworking foreground detail")
     _expect("front-right quarter view" in str(pair.get("softcore_camera_directive")), "soft pair camera directive was not preserved")
     _expect("back-right quarter view" in str(pair.get("hardcore_camera_directive")), "hard pair camera directive was not preserved")
     soft_row = pair.get("softcore_row") or {}
@@ -6065,9 +6067,19 @@ def smoke_pov_camera_scene() -> None:
     scene_directive = _expect_text("pov_camera_scene.hard_camera_scene", hard_row.get("camera_scene_directive"), 40)
     _expect("from POV" in scene_directive, "POV camera scene should be marked as first-person")
     _expect("not in the lower foreground" in scene_directive, "POV camera scene should keep location anchors out of lower foreground")
+    _expect(
+        "near desk edge" not in scene_directive and "laptop corner" not in scene_directive and "chair back" not in scene_directive,
+        "POV camera scene should not reuse coworking foreground anchors as viewer-side objects",
+    )
+    _expect(
+        "lower foreground is reserved for POV body or hand cues" in scene_directive,
+        "POV camera scene should reserve the lower foreground for first-person body cues",
+    )
     krea = krea_formatter.format_krea2_prompt("", metadata_json=_json(pair), target="hardcore")
     prompt = krea.get("krea_prompt") or ""
     _expect("from POV" in prompt, "Krea POV prompt lost camera-scene directive")
+    _expect("laptop corner" not in prompt, "Krea POV prompt leaked foreground desk anchor into camera scene")
+    _expect("lower foreground is reserved for POV body or hand cues" in prompt, "Krea POV prompt lost POV foreground reservation")
     _expect("Camera:" not in prompt, "Krea POV prompt should not emit normal third-person camera directive")