Expand environment image prompt details

AGENTS.md

@@ -150,6 +150,10 @@ When the user says to prepare for imminent compaction:
   fictional objects and body affordances in visible terms: geometry, materials,
   lighting, stance, interfaces, tools, supports, cables, fluids, screens, and
   crowd positions.
+- Do the same for environments. A location label like "manifold line," "cavity
+  yard," "supervisor glass," or "anchor rail" is not a drawable scene until the
+  prompt describes its visible geometry, materials, interfaces, lights, and
+  scale.
 - If a fixture has a target visual style, store that style cue once in the
   sidecar and include it in every assembled image prompt. Do not rely on memory
   or unstated website context to carry the look.

docs/architecture/ink-branching-scenes.md

@@ -420,6 +420,13 @@ interface, cavity seal assembly, or life-support bypass kit, describe the
 shape, scale, materials, lights, cables, seals, or displays that make it
 drawable.
 
+The same rule applies to environments. "Manifold line," "cavity yard,"
+"supervisor glass," and "anchor rail" are not images by themselves. A base
+prompt should describe the physical space a general image model can draw:
+corridor shape, wall/ceiling/floor geometry, pipe runs, cable bundles, panels,
+screens, rails, doors, windows, gauges, warning lights, materials, scale, and
+where bodies stand inside that geometry.
+
 Every visual section should identify:
 
 - `visual_scene_id`: stable handle for the screen or key frame

examples/ink/pallas-species-strikes.branch-and-fold.v0.training.json

@@ -602,7 +602,7 @@
       {
         "visual_scene_id": "pallas_nara_manifold_walk",
         "ink_anchor": "walk_with_nara",
-        "base_image_prompt": "Close medium shot of Lio walking beside Nara-7 along the manifold line: overlapping patch plates on the wall look like metal scars, amber fault diagrams glow on Lio's slate, Nara's numbered batch tag is visible but not dominant, and her fingers hover near the wall as if listening through touch.",
+        "base_image_prompt": "Close medium shot inside Service Ring Kappa, a narrow industrial maintenance corridor built against a curved asteroid-rock wall: the manifold line is a waist-high run of ribbed bronze-gray pipe trunks, bundled fiber cables, bolted valve housings, removable patch plates, small amber route-diagram screens, pressure gauges, warning glyphs, and hand-sized diagnostic touch panels mounted along the wall like a long mechanical spine. Scuffed deck plates with yellow-white safety stripes curve underfoot; overhead work lights cast hard amber reflections through faint industrial mist. Lio walks beside Nara-7 along this pipe-and-panel service wall; amber fault diagrams glow on Lio's slate, Nara's numbered batch tag is visible but not dominant, and her fingers hover near a ribbed access plate as if listening through touch.",
         "visible_characters": [
           {
             "agent_id": "lio_vale",
@@ -1420,7 +1420,8 @@
   "branch_compiler_notes": {
     "review_notes": [
       "Added a global Media Molecule Dreams-style volumetric brushstroke cue so image prompt assembly carries the intended non-photoreal handmade visual language without repeating it in every scene.",
-      "Corrected image prompt assembly: visual_continuity_notes are non-prompt tooling constraints, and assembled prompts include only character refs for characters actually visible in the frame."
+      "Corrected image prompt assembly: visual_continuity_notes are non-prompt tooling constraints, and assembled prompts include only character refs for characters actually visible in the frame.",
+      "Expanded pallas_nara_manifold_walk so the environment is self-contained; manifold line is described as visible pipe trunks, cable bundles, valve housings, patch plates, screens, gauges, and touch panels rather than assumed lore."
     ]
   }
 }

prompts/README.md

@@ -46,5 +46,8 @@ should be authored here first and referenced from artifacts or tools.
 - Treat continuity notes as tooling constraints, not prompt text. If the image
   model needs a continuity fact, put it in visible terms inside the base prompt,
   character ref, or modifier.
+- Treat environment labels as handles, not images. A prompt that names a
+  location or fixture must also describe its visible geometry, materials,
+  lights, interfaces, and scale.
 - If a prompt is used for a capture, preserve the exact rendered prompt in the
   capture and include the prompt file path when the capture schema allows it.

prompts/visual-scene-continuity-reviewer.md

@@ -51,6 +51,9 @@ intent, but they do not excuse missing reader/player-facing structure.
   atmosphere, materials, visible interfaces, bodies, tools, and important props?
 - Are specialized terms translated into visible forms? For example, explain
   what a wet-service cradle looks like, not just what it is called.
+- Are environment labels translated into visible forms? For example, explain
+  what a manifold line, cavity yard, supervisor glass, or anchor rail looks
+  like instead of treating the label as a picture.
 - Does the art plan distinguish wide establishing shots, medium character
   staging, object closeups, and alert/climax frames when the prose focus shifts?
 
@@ -150,6 +153,9 @@ Each finding must include:
   assembly. A beautiful scene prompt in the wrong visual language is still the
   wrong prompt.
 - Do not accept visual prompts that rely on unexplained setting jargon.
+- Do not accept environment labels as visual descriptions. "Along the manifold
+  line" is not enough unless the prompt describes the pipe/cable/panel/valve
+  structure and surrounding corridor geometry.
 - Do not accept named characters as visual descriptions. "Lio Vale stands by
   the manifold" is not enough unless the sidecar provides a stable visual
   reference for Lio and the prompt assembly path says to include it.
@@ -182,6 +188,8 @@ Use these labels when applicable:
 - `pivotal_moment_not_segmented`
 - `single_prompt_for_multi_scene_fixture`
 - `image_prompt_lore_password`
+- `environment_lore_password`
+- `weak_environment_geometry`
 - `weak_geometry_prompt`
 - `missing_global_style_cue`
 - `style_cue_not_in_prompt_assembly`

state/evidence.jsonl

@@ -47,3 +47,4 @@
 {"ts": "2026-05-03T09:25:38+00:00", "type": "review", "status": "ok", "note": "Added stable visual character refs for the Pallas illustrated IF fixture and updated visual review doctrine so named characters cannot rely on imagegen knowing their appearance from names alone."}
 {"ts": "2026-05-03T09:30:46+00:00", "type": "design", "status": "ok", "note": "Illustrated IF fixtures with a target look should store a global visual style cue and include it in every assembled image prompt; the Pallas fixture now carries the Dreams-style volumetric brushstroke cue."}
 {"ts": "2026-05-03T09:37:16+00:00", "type": "design", "status": "ok", "note": "Corrected illustrated prompt assembly: prompts should include only in-frame character refs, while continuity notes remain tooling constraints unless rewritten as concrete visible prompt text."}
+{"ts": "2026-05-03T09:39:31+00:00", "type": "design", "status": "ok", "note": "Corrected illustrated prompt doctrine: environment labels such as manifold line must be rewritten as concrete visible geometry, materials, interfaces, lights, and scale before reaching imagegen."}