v4 execution lock: creator evidence

.playwright-mcp/page-2026-06-12T12-58-04-245Z.yml

@@ -0,0 +1,12 @@
+- generic [active] [ref=e1]:
+  - region "Notifications alt+T"
+  - alert [ref=e7]
+  - dialog "Privacy" [ref=e9]:
+    - generic [ref=e10]:
+      - generic [ref=e13]: By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts.
+      - generic [ref=e15]:
+        - button "Cookies Settings, Opens the preference center dialog" [ref=e16] [cursor=pointer]: Cookies Settings
+        - generic [ref=e17]:
+          - button "Reject All" [ref=e18] [cursor=pointer]
+          - button "Accept All Cookies" [ref=e19] [cursor=pointer]
+      - button "Close" [ref=e21] [cursor=pointer]

.playwright-mcp/page-2026-06-12T21-30-33-874Z.yml

@@ -0,0 +1,3 @@
+- generic [ref=e5]:
+  - img [ref=e9]
+  - paragraph [ref=e20]: Loading...

.playwright-mcp/page-2026-06-12T21-31-39-288Z.yml

@@ -0,0 +1,5 @@
+- generic [ref=e4]:
+  - generic [ref=e5]:
+    - img [ref=e9]
+    - paragraph [ref=e20]: Loading...
+  - main [ref=e22]

.playwright-mcp/page-2026-06-13T05-01-53-264Z.yml

@@ -0,0 +1,3 @@
+- generic [ref=e5]:
+  - img [ref=e9]
+  - paragraph [ref=e20]: Loading...

.playwright-mcp/page-2026-06-13T14-11-22-087Z.yml

@@ -0,0 +1,5 @@
+- generic [ref=e4]:
+  - generic [ref=e5]:
+    - img [ref=e9]
+    - paragraph [ref=e20]: Loading...
+  - main [ref=e22]

.playwright-mcp/page-2026-06-13T14-19-01-146Z.yml

@@ -0,0 +1,3 @@
+- generic [ref=e5]:
+  - img [ref=e9]
+  - paragraph [ref=e20]: Loading...

README.md

@@ -12,14 +12,9 @@ short_description: Governed gothic couture visual creation command center
 models:
   - black-forest-labs/FLUX.2-klein-9B
   - black-forest-labs/FLUX.2-klein-4B
-  - Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf
   - nvidia/LocateAnything-3B
   - openbmb/MiniCPM-V-4.6
   - nvidia/NVIDIA-Nemotron-Parse-v1.2
-  - openbmb/MiniCPM5-1B
-  - onnx-community/functiongemma-270m-it-ONNX
-  - hexgrad/Kokoro-82M
-  - netflix/void-model
 tags:
   - gradio
   - mcp-server
@@ -30,8 +25,6 @@ tags:
   - best-agent
   - best-demo
   - openbmb
-  - nvidia
-  - modal
   - codex
 ---
 
@@ -57,17 +50,14 @@ The interface is built around a command-center surface:
 
 Pinned lanes do not rotate:
 
-- `image_generation`: Raven Quality Stack with `black-forest-labs/FLUX.2-klein-9B` as the flagship image/edit lane
+- `image_generation`: Raven Quality FLUX.2 Klein 9B image lane, with FLUX.2 Klein 4B as Tiny Titan sidecar/fallback
 - `grounding`: NVIDIA LocateAnything-3B grounding anchor
 - `security`: ST3GG defensive scanner/export gate
 
 Sponsor/evidence lanes are optional but first-class when secrets are configured:
 
-- `Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf` (11.91B): quality/taste/lore critique lane for private or configured judge use.
 - `openbmb/MiniCPM-V-4.6` (1.30B): visual judge for wardrobe, footwear, material drift, lore continuity, and export notes.
 - `nvidia/NVIDIA-Nemotron-Parse-v1.2` (0.94B): structured evidence/parser lane for NVIDIA/Nemotron claim support.
-- `hexgrad/Kokoro-82M` (0.082B): optional lore narration lane.
-- `netflix/void-model` (5B CogVideoX-based): Modal/offline video repair sample lane, not a blocking Space runtime default.
 
 Helper lanes may rotate with quota, license, health, and parameter-budget checks:
 
@@ -78,12 +68,12 @@ Helper lanes may rotate with quota, license, health, and parameter-budget checks
 - HF catalog research
 - Modal job runner
 
-The default preset is **Raven Quality Stack**. `black-forest-labs/FLUX.2-klein-4B` remains available only as a Tiny Titan/public-safe sidecar and fallback if the gated 9B lane is unavailable. OFFELLIA heretic/obliterated-style variants stay private research only and never disable consent, provenance, ST3GG, export, or dataset-partition gates.
+Public demo mode uses the Raven Quality FLUX.2 Klein 9B lane when Space access is configured, while preserving the 4B sidecar for fallback/Tiny Titan evidence. Private research mode can add OFFELLIA/Gemma routes, but it never disables consent, provenance, ST3GG, export, or dataset-partition gates.
 
 ## Current Features
 
 - Gradio Blocks dashboard with split update regions.
-- Real FLUX.2 Klein 9B-first image generation on Hugging Face ZeroGPU when runtime access is configured, with an honest 4B sidecar fallback.
+- Real FLUX.2 Klein 9B-first image generation on Hugging Face ZeroGPU when runtime access is configured, with honest 4B sidecar fallback.
 - Above-fold ST3GG trust strip with safe-vs-blocked fixture evidence.
 - Generated artifact ST3GG scan and checkpoint/export state.
 - Optional MiniCPM-V and Nemotron provider evidence lanes with explicit configured/missing-secret status.
@@ -100,18 +90,25 @@ The default preset is **Raven Quality Stack**. `black-forest-labs/FLUX.2-klein-4
 | Target | Evidence status |
 | --- | --- |
 | Gradio Space | App runs as a public Hugging Face Gradio Space with `mcp_server=True`. |
-| <=32B models | Raven Quality Stack is 28.50B: FLUX.2 Klein 9B + OFFELLIA Q4 Gemma 12B + LocateAnything 3.83B + MiniCPM-V 1.30B + Nemotron Parse 0.94B + MiniCPM5 1.08B + FunctionGemma 0.27B + Kokoro 0.082B. |
+| <=32B models | Raven Quality stack is 16.42B: FLUX.2 Klein 9B + LocateAnything 3.83B + MiniCPM-V 1.30B + Nemotron Parse 0.94B + MiniCPM5 1.08B + FunctionGemma 0.27B. |
 | Off Brand | Custom command-center UI, dense inspector, workflow graph, wardrobe/lore drawer, and provider cards. |
 | Best Agent | Multi-step prompt, generation, scan, judge, checkpoint, export workflow. |
 | OpenBMB | Claimed only when MiniCPM-V returns success status in an export packet. |
 | NVIDIA | Claimed only when Nemotron returns success status in an export packet. LocateAnything remains visible but is not the Nemotron claim by itself. |
-| Modal | Sidecar-only until a real Modal job is documented; target lane is `netflix/void-model` video repair. |
 | OpenAI Codex | Development branch and PR include Codex-authored implementation commits. |
 | Demo / social | Add final links here before submission: `DEMO_VIDEO_URL` and `SOCIAL_POST_URL`. |
 
-Tiny Titan is not the flagship story. It can be claimed only from a successful sidecar export packet where every active sidecar model is <=4B.
+Tiny Titan can be claimed only from a successful sidecar export packet using the 4B mode. It must not weaken the main Raven Quality stack.
 
-See [docs/SUBMISSION_ASSETS.md](docs/SUBMISSION_ASSETS.md) for the demo script, social post draft, thumbnail direction, and final link checklist.
+## Merit Badges
+
+| Badge | Bonus Quest | Status | Evidence |
+| --- | --- | --- | --- |
+| `off-brand` | Off-Brand | Ready | Gothic couture command-center UI with workflow graph, inspector, wardrobe/lore drawer, and ST3GG trust strip. |
+| `best-agent` | Best Agent | Ready when export packet exists | Multi-step weave: prompt refinement, generation, scan, optional judges, checkpoint, and governed export. |
+| `best-demo` | Best Demo | Pending final asset | Requires final demo video link and social post link before submission. |
+| `openbmb` | OpenBMB sponsor | Conditional | Claimed only when MiniCPM-V returns `success` in an export packet. |
+| `nvidia` | NVIDIA/Nemotron sponsor | Conditional | Claimed only when Nemotron returns `success`; LocateAnything remains grounding evidence, not a Nemotron claim. |
 
 ## Local Setup
 

app.py

@@ -4,8 +4,10 @@ from __future__ import annotations
 
 import os
 import sys
+import hashlib
 from pathlib import Path
 from typing import Any
+from urllib.parse import urlparse
 
 import gradio as gr
 
@@ -49,32 +51,10 @@ MODEL_RELAY = WeaverModelRelay()
 
 def _default_operator_state() -> dict[str, Any]:
     return {
-        "active_preset": "Raven Quality Stack",
         "provider_state": "idle",
         "checkpoint": "pending",
         "export": "pending",
         "message": "No operator action yet.",
-        "modal_video_repair": {
-            "status": "deferred",
-            "repo_id": "netflix/void-model",
-            "provider": "modal",
-            "message": "Offline/Modal VOID repair sample is documented but not a blocking runtime default.",
-        },
-        "offellia_judge": {
-            "status": "deferred_local",
-            "repo_id": "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf",
-            "message": "Quality/taste judge lane is declared in the Raven stack and runs only when local GGUF runtime is configured.",
-        },
-        "audio_lore_tts": {
-            "status": "optional",
-            "repo_id": "hexgrad/Kokoro-82M",
-            "message": "Lore narration lane is off by default for the public demo.",
-        },
-        "tiny_titan_sidecar": {
-            "status": "available",
-            "repo_id": "black-forest-labs/FLUX.2-klein-4B",
-            "message": "Public-safe 4B sidecar remains selectable without weakening the quality preset.",
-        },
     }
 
 
@@ -83,7 +63,7 @@ def _zero_gpu_entrypoint(fn: Any) -> Any:
     gpu_decorator = getattr(spaces, "GPU", None) if spaces is not None else None
     if gpu_decorator is None:
         return fn
-    return gpu_decorator(duration=900)(fn)
+    return gpu_decorator(duration=300)(fn)
 
 
 def _relay_snapshot(adult_mode: bool = False) -> dict[str, Any]:
@@ -99,6 +79,103 @@ def _file_path(uploaded: Any) -> str | None:
     return str(path) if path else None
 
 
+def _safe_file_hash(path: str | None) -> tuple[str | None, int | None]:
+    if not path:
+        return None, None
+    try:
+        target = Path(path)
+        data = target.read_bytes()
+    except OSError:
+        return None, None
+    return hashlib.sha256(data).hexdigest(), len(data)
+
+
+def _safe_reference_url_metadata(reference_url: str | None) -> dict[str, Any] | None:
+    if not reference_url:
+        return None
+    parsed = urlparse(reference_url.strip())
+    if parsed.scheme not in {"http", "https"} or not parsed.netloc:
+        return {"source": "url", "status": "invalid_url", "message": "Reference URL must be http(s)."}
+    url_hash = hashlib.sha256(reference_url.strip().encode("utf-8")).hexdigest()
+    return {
+        "source": "url",
+        "status": "metadata_only",
+        "domain": parsed.netloc.lower(),
+        "url_hash": url_hash,
+        "message": "URL stored as metadata only; Space runtime does not crawl or copy shop images.",
+    }
+
+
+def _reference_metadata(uploaded: Any, reference_url: str | None, scan: dict[str, Any]) -> list[dict[str, Any]]:
+    records: list[dict[str, Any]] = []
+    path = _file_path(uploaded)
+    if path:
+        file_hash, size = _safe_file_hash(path)
+        records.append(
+            {
+                "source": "upload",
+                "basename": Path(path).name,
+                "sha256": file_hash,
+                "size_bytes": size,
+                "st3gg_status": scan.get("status"),
+                "export_gate": scan.get("export_gate"),
+                "magic": scan.get("magic"),
+                "extension": scan.get("extension"),
+            }
+        )
+    url_record = _safe_reference_url_metadata(reference_url)
+    if url_record:
+        records.append(url_record)
+    return records
+
+
+def _creator_controls(
+    reasoning_mode: str,
+    video_preset: str,
+    silhouette: str | None = None,
+    outerwear: str | None = None,
+    upper_body: str | None = None,
+    footwear: str | None = None,
+    palette: str | None = None,
+    hardware: str | None = None,
+    locate_focus: list[str] | None = None,
+) -> dict[str, Any]:
+    wardrobe = {
+        "silhouette": silhouette or "structured long coat",
+        "outerwear": outerwear or "black patent leather long coat",
+        "upper_body": upper_body or "Chantilly lace neckline",
+        "footwear": footwear or "platform boots",
+        "palette": palette or "black, crimson, cyan neon",
+        "hardware": hardware or "crimson hardware",
+        "locked_slots": ["outerwear", "upper_body", "footwear", "jewelry"],
+        "locate_focus": locate_focus or ["outerwear", "footwear", "jewelry"],
+    }
+    return {
+        "reasoning_mode": reasoning_mode,
+        "video_preset": video_preset,
+        "wardrobe": wardrobe,
+        "generation": {
+            "flux_primary": "black-forest-labs/FLUX.2-klein-9B",
+            "flux_sidecar": "black-forest-labs/FLUX.2-klein-4B",
+            "lora_policy": "attempt compatible runtime adapter; report loaded/skipped/failed",
+        },
+    }
+
+
+def _prompt_with_controls(prompt: str, controls: dict[str, Any]) -> str:
+    wardrobe = controls.get("wardrobe", {})
+    additions = [
+        wardrobe.get("silhouette"),
+        wardrobe.get("outerwear"),
+        wardrobe.get("upper_body"),
+        wardrobe.get("footwear"),
+        wardrobe.get("palette"),
+        wardrobe.get("hardware"),
+    ]
+    suffix = ", ".join(str(item) for item in additions if item)
+    return f"{prompt}\nWardrobe controls: {suffix}" if suffix else prompt
+
+
 def _generated_output_path(operator_state: dict[str, Any] | None) -> str | None:
     generation = (operator_state or {}).get("generation") or {}
     output_path = generation.get("output_path")
@@ -159,16 +236,41 @@ def run_weave(
     adult_mode: bool,
     upload: Any,
     active_section: str,
+    silhouette: str | None = None,
+    outerwear: str | None = None,
+    upper_body: str | None = None,
+    footwear: str | None = None,
+    palette: str | None = None,
+    hardware: str | None = None,
+    reference_url: str | None = None,
 ) -> tuple[Any, ...]:
     prompt = prompt.strip() or DEFAULT_PROMPT
+    controls = _creator_controls(
+        reasoning_mode=reasoning_mode,
+        video_preset=video_preset,
+        silhouette=silhouette,
+        outerwear=outerwear,
+        upper_body=upper_body,
+        footwear=footwear,
+        palette=palette,
+        hardware=hardware,
+    )
+    controlled_prompt = _prompt_with_controls(prompt, controls)
+    reference_scan = scan_file(_file_path(upload))
+    reference_metadata = _reference_metadata(upload, reference_url, reference_scan)
     run = build_command_center_run(
-        prompt=prompt,
+        prompt=controlled_prompt,
         mode=reasoning_mode,
         video_preset=video_preset,
         adult_mode=adult_mode,
+        creator_controls=controls,
+        reference_metadata=reference_metadata,
+    )
+    generation = generate_flux_image(
+        run.refined_prompt.refined,
+        seed=_checkpoint_seed(run.checkpoint.checkpoint_id),
+        adult_mode=adult_mode,
     )
-    reference_scan = scan_file(_file_path(upload))
-    generation = generate_flux_image(run.refined_prompt.refined, seed=_checkpoint_seed(run.checkpoint.checkpoint_id))
     generated_scan = scan_file(generation.output_path) if generation.output_path else scan_file(None)
     minicpm = judge_with_minicpm(
         prompt=run.refined_prompt.refined,
@@ -185,31 +287,17 @@ def run_weave(
     if generation.status == "success":
         provider_state = "generated"
     operator_state = {
-        **_default_operator_state(),
-        "active_preset": "Raven Quality Stack",
         "provider_state": provider_state,
         "checkpoint": "pending_review",
         "export": generated_scan.get("export_gate", "pending"),
         "message": generation.message or "Run packet generated. Human checkpoint required before provider promotion or export.",
         "generation": generation.to_dict(),
+        "creator_controls": controls,
+        "reference_metadata": reference_metadata,
         "reference_scan": reference_scan,
         "generated_scan": generated_scan,
         "minicpm_judge": minicpm.to_dict(),
         "nemotron_evidence": nemotron.to_dict(),
-        "locateanything_grounding": {
-            "status": run.inspection.status,
-            "repo_id": run.inspection.locate_model,
-            "targets": [
-                {
-                    "slot_name": target.slot_name,
-                    "query": target.query,
-                    "expected_region": target.expected_region,
-                    "confidence": target.confidence,
-                }
-                for target in run.inspection.targets[:6]
-            ],
-            "drift_flags": run.inspection.drift_flags,
-        },
     }
     regions = _dashboard_regions(
         run=run,
@@ -329,10 +417,12 @@ def scan_reference(
     upload: Any,
     active_section: str,
     operator_state: dict[str, Any] | None,
+    reference_url: str | None = None,
 ) -> tuple[Any, ...]:
     state = operator_state or _default_operator_state()
     reference_path = _file_path(upload)
     reference_scan = scan_file(reference_path)
+    reference_metadata = _reference_metadata(upload, reference_url, reference_scan)
     generated_scan = _authoritative_generated_scan(state)
     minicpm = None
     if run is not None and reference_path:
@@ -342,27 +432,10 @@ def scan_reference(
             scan=reference_scan,
             wardrobe_summary=_wardrobe_summary(run),
         )
-    locate_plan = {}
-    if run is not None:
-        locate_plan = {
-            "status": run.inspection.status,
-            "repo_id": run.inspection.locate_model,
-            "source": "reference_scan",
-            "targets": [
-                {
-                    "slot_name": target.slot_name,
-                    "query": target.query,
-                    "expected_region": target.expected_region,
-                    "confidence": target.confidence,
-                }
-                for target in run.inspection.targets[:6]
-            ],
-            "drift_flags": run.inspection.drift_flags,
-        }
     next_state = {
         **state,
         **({"reference_judge": minicpm.to_dict()} if minicpm else {}),
-        **({"reference_locate_plan": locate_plan, "locateanything_grounding": locate_plan} if locate_plan else {}),
+        "reference_metadata": reference_metadata,
         "reference_scan": reference_scan,
         "reference_export_gate": reference_scan.get("export_gate", "pending"),
         "export": state.get("export", generated_scan.get("export_gate", "pending")),
@@ -413,36 +486,25 @@ def export_packet(
     scan: dict[str, Any] | None,
     active_section: str,
     operator_state: dict[str, Any] | None,
-    override_reason: str = "",
 ) -> tuple[Any, ...]:
     state = operator_state or _default_operator_state()
     scan = _authoritative_generated_scan(state)
-    override_reason = (override_reason or "").strip()
     if run is None:
         next_state = {**state, "provider_state": "blocked", "export": "blocked", "message": "Export blocked: no active run packet exists."}
     elif state.get("checkpoint") != "approved":
         next_state = {**state, "provider_state": "blocked", "export": "blocked", "message": "Export blocked: human checkpoint has not been approved."}
     elif not _generated_output_path(state):
         next_state = {**state, "provider_state": "blocked", "export": "blocked", "message": "Export blocked: no generated artifact exists."}
-    elif scan.get("export_gate") != "clear" and not override_reason:
-        next_state = {**state, "provider_state": "blocked", "export": scan.get("export_gate", "blocked"), "message": "Export blocked: ST3GG gate is not clear. Add an explicit override reason to prepare a reviewed evidence packet."}
+    elif scan.get("export_gate") != "clear":
+        next_state = {**state, "provider_state": "blocked", "export": scan.get("export_gate", "blocked"), "message": "Export blocked: ST3GG gate is not clear."}
     else:
-        export_state = "clear" if scan.get("export_gate") == "clear" else "override"
-        export_input_state = {
-            **state,
-            **({"st3gg_override_reason": override_reason} if override_reason and scan.get("export_gate") != "clear" else {}),
-        }
-        export = write_export_packet(run=run, scan=scan, operator_state=export_input_state, adult_mode=adult_mode)
+        export = write_export_packet(run=run, scan=scan, operator_state=state, adult_mode=adult_mode)
         next_state = {
-            **export_input_state,
+            **state,
             "provider_state": "exported",
-            "export": export_state,
+            "export": "clear",
             "export_packet": {"path": export["path"]},
-            "message": (
-                f"Governed export packet prepared with ST3GG override recorded: {export['path']}"
-                if export_state == "override"
-                else f"Governed export packet prepared: {export['path']}"
-            ),
+            "message": f"Governed export packet prepared: {export['path']}",
         }
     return _render_stateful(run, adult_mode, scan, active_section, next_state)
 
@@ -521,6 +583,43 @@ with gr.Blocks(title="NEXUS Visual Weaver") as demo:
                     value="Wan2.2 I2V",
                     label="Video Path Preset",
                 )
+        with gr.Row():
+            silhouette = gr.Dropdown(
+                ["structured long coat", "fitted gothic bodice", "layered tactical silhouette"],
+                value="structured long coat",
+                label="Silhouette",
+            )
+            outerwear = gr.Dropdown(
+                ["black patent leather long coat", "faux fur collar coat", "tailored rain slicker"],
+                value="black patent leather long coat",
+                label="Outerwear",
+            )
+            upper_body = gr.Dropdown(
+                ["Chantilly lace neckline", "black mesh layer", "structured corset bodice"],
+                value="Chantilly lace neckline",
+                label="Upper Body",
+            )
+            footwear = gr.Dropdown(
+                ["platform boots", "patent leather heels", "armored couture boots"],
+                value="platform boots",
+                label="Footwear",
+            )
+        with gr.Row():
+            palette = gr.Dropdown(
+                ["black, crimson, cyan neon", "obsidian, pearl, crimson", "graphite, magenta, cold blue"],
+                value="black, crimson, cyan neon",
+                label="Palette",
+            )
+            hardware = gr.Dropdown(
+                ["crimson hardware", "silver occult buckles", "holographic NEXUS sigils"],
+                value="crimson hardware",
+                label="Hardware",
+            )
+            reference_url = gr.Textbox(
+                label="Reference URL (metadata only)",
+                placeholder="https://shop.example/reference-garment",
+                scale=2,
+            )
         with gr.Row():
             upload = gr.File(
                 label="Reference / Output For ST3GG Scan",
@@ -536,13 +635,6 @@ with gr.Blocks(title="NEXUS Visual Weaver") as demo:
             )
             run_btn = gr.Button("Run Active Weave", variant="primary", scale=1)
             stop_btn = gr.Button("Stop Provider Job", variant="stop", interactive=False, scale=1)
-        override_reason = gr.Textbox(
-            value="",
-            label="ST3GG Override Reason",
-            placeholder="Required only to export a reviewed/blocked artifact. Explain why the evidence packet may be written.",
-            lines=2,
-            max_lines=3,
-        )
         with gr.Row(elem_id="nw-operator-actions", elem_classes=["nw-operator-actions"]):
             scan_btn = gr.Button("Scan Reference", scale=1)
             checkpoint_btn = gr.Button("Approve Checkpoint", scale=1)
@@ -596,17 +688,21 @@ with gr.Blocks(title="NEXUS Visual Weaver") as demo:
 
     operator_outputs = dashboard_outputs + [operator_state, stop_btn]
 
-    run_btn.click(
+    run_click = run_btn.click(
         fn=run_weave,
-        inputs=[prompt, reasoning_mode, video_preset, adult_mode, upload, section_nav],
+        inputs=[prompt, reasoning_mode, video_preset, adult_mode, upload, section_nav, silhouette, outerwear, upper_body, footwear, palette, hardware, reference_url],
         outputs=stateful_outputs,
         api_name="run_active_weave",
+        concurrency_limit=1,
+        concurrency_id="flux-gpu",
     )
-    prompt.submit(
+    run_submit = prompt.submit(
         fn=run_weave,
-        inputs=[prompt, reasoning_mode, video_preset, adult_mode, upload, section_nav],
+        inputs=[prompt, reasoning_mode, video_preset, adult_mode, upload, section_nav, silhouette, outerwear, upper_body, footwear, palette, hardware, reference_url],
         outputs=stateful_outputs,
         api_name=False,
+        concurrency_limit=1,
+        concurrency_id="flux-gpu",
     )
     adult_mode.change(
         fn=toggle_adult_visibility,
@@ -624,42 +720,51 @@ with gr.Blocks(title="NEXUS Visual Weaver") as demo:
             operator_state,
         ],
         api_name="toggle_adult_catalog",
+        queue=False,
     )
     section_nav.change(
         fn=refresh_section,
         inputs=[section_nav, adult_mode, active_run_state, scan_state, operator_state],
         outputs=[command_rail_html, operations_html, inspector_html, artifact_html, provider_html, scan_json],
         api_name=False,
+        queue=False,
     )
     scan_btn.click(
         fn=scan_reference,
-        inputs=[active_run_state, adult_mode, upload, section_nav, operator_state],
+        inputs=[active_run_state, adult_mode, upload, section_nav, operator_state, reference_url],
         outputs=dashboard_outputs + [operator_state, stop_btn, scan_state],
         api_name="scan_reference",
+        queue=False,
     )
     checkpoint_btn.click(
         fn=approve_checkpoint,
         inputs=[active_run_state, adult_mode, scan_state, section_nav, operator_state],
         outputs=operator_outputs,
         api_name="approve_checkpoint",
+        queue=False,
     )
     export_btn.click(
         fn=export_packet,
-        inputs=[active_run_state, adult_mode, scan_state, section_nav, operator_state, override_reason],
+        inputs=[active_run_state, adult_mode, scan_state, section_nav, operator_state],
         outputs=operator_outputs,
         api_name="prepare_export_packet",
+        queue=False,
     )
     stop_btn.click(
         fn=stop_provider_job,
         inputs=[active_run_state, adult_mode, scan_state, section_nav, operator_state],
         outputs=operator_outputs,
         api_name="stop_provider_job",
+        queue=False,
+        cancels=[run_click, run_submit],
     )
     reset_btn.click(
         fn=reset_demo,
         inputs=[adult_mode, section_nav],
         outputs=stateful_outputs,
         api_name="reset_demo_state",
+        queue=False,
+        cancels=[run_click, run_submit],
     )
     demo.load(
         fn=lambda: (render_catalog_table(False), catalog_summary(False), scan_file(None), scan_file(None), _default_operator_state()),

docs/HACKATHON_EVALUATION.md

@@ -13,11 +13,11 @@ NEXUS Visual Weaver should open as a working command center, not a landing page.
 ## Current Strengths
 
 - Gradio-compatible app shape with `mcp_server=True`.
-- Pinned model governance is visible: FLUX.2 Klein 9B, LocateAnything-3B, and ST3GG.
-- Real FLUX.2 Klein 9B-first generation is wired for HF Space and falls back to an honest 4B Tiny Titan sidecar when the gated lane is unavailable.
+- Pinned model governance is visible: FLUX.2 Klein 4B, LocateAnything-3B, and ST3GG.
+- Real FLUX.2 Klein 4B generation is wired for HF Space and falls back to an honest dry-run state outside Space.
 - Generated artifacts are scanned by ST3GG before checkpoint/export.
 - Above-fold trust strip makes ST3GG verdict, export gate, fixture evidence, and adult-mode safety boundaries visible immediately.
-- OpenBMB MiniCPM-V 4.6, NVIDIA Nemotron, OFFELLIA Q4, LocateAnything, Kokoro TTS, and Modal VOID evidence lanes are represented with missing-secret/deferred/failed/success states.
+- OpenBMB MiniCPM-V 4.6 and NVIDIA Nemotron evidence lanes are represented as real optional provider adapters with missing-secret/failed/success states.
 - Adult Mode starts off and is framed as catalog scope, not a safety bypass.
 - ModelRelay/GMR helper rotation is represented without replacing pinned lanes.
 - Tests cover catalog scope, workflow planning, ModelRelay behavior, scanner evidence, and dashboard fallback rendering.
@@ -32,21 +32,21 @@ NEXUS Visual Weaver should open as a working command center, not a landing page.
 
 ## Next Implementation Priority
 
-1. Keep Raven Quality Stack as the submission narrative; use Tiny Titan only as a sidecar export.
-2. Configure OpenBMB and Nemotron Space secrets if sponsor prize claims are desired.
-3. Run one live Space weave and prepare an export packet.
-4. Run/document one Modal sidecar job only if it can complete without risking the main Space.
-5. Capture demo video, create social post, and add final links to README.
+1. Configure OpenBMB and Nemotron Space secrets if prize claims are desired.
+2. Run one live Space weave and prepare an export packet.
+3. Capture demo video and create social post.
+4. Add final demo/social URLs to README.
+5. Add Playwright/browser visual checks for desktop and mobile overflow once CI is unblocked.
 
 ## Prize Claim Evidence Rules
 
 | Prize or badge | Current stance |
 | --- | --- |
-| Build Small base eligibility | Gradio Space, each active model <32B, and public app path are ready; demo/social links still required. |
+| Build Small base eligibility | Gradio Space, <=32B stack, and public app path are ready; demo/social links still required. |
 | Off Brand | Strong custom command-center UI signal. |
 | Best Agent | Multi-step governed workflow is implemented through callbacks and export packet. |
 | OpenBMB | Claim only after MiniCPM-V returns `success` in export evidence. |
 | NVIDIA | Claim only after Nemotron returns `success` in export evidence. |
 | OpenAI Codex | GitHub branch/PR provides Codex development trail. |
-| Tiny Titan | Sidecar-only: claim only from an export packet where every active sidecar model is <=4B. |
-| Modal | Not claimed unless a real `netflix/void-model` or equivalent Modal job runs and is documented. |
+| Tiny Titan | Public demo stack is eligible: active public models are each <=4B. |
+| Modal | Not claimed unless a real Modal job runs. |

docs/HANDOFF_FINAL_HACKATHON.md

@@ -3,30 +3,24 @@
 ## Current Target
 
 - GitHub repo: `https://github.com/specimba/NEXUS-Visual-Weaver`
-- Working branch: `main`
+- Working branch: `codex/specimba/ui-polish-command-center`
 - HF Space: `build-small-hackathon/NEXUS_Visual_Weaver`
 - Public Space URL: `https://build-small-hackathon-nexus-visual-weaver-a107340.hf.space/`
 - HF rollback SHA: `410a467c55d11e7308249198bd5fe0b2c190aec6`.
-- Stable rollback tag: `stable-raven-quality-20260615` -> `aa37d4d`.
-- Last verified code commit: `aa37d4d` (`fix: load flux2 klein pipeline`).
-- Last verified HF Space commit: `effbdb4c6a1ea98410165310caea10097ffa0ca6`.
-- Last verified HF runtime: `RUNNING`, ZeroGPU, dev mode disabled, `/data` bucket mounted.
 - Branch discipline: use only `main` and `codex/specimba/ui-polish-command-center`; no extra recovery branches.
-- Primary goal: finish a countable Build Small submission with Raven Quality Stack generation, ST3GG scan, LocateAnything grounding, optional OpenBMB MiniCPM-V judge evidence, optional NVIDIA Nemotron evidence, optional Modal VOID sidecar evidence, checkpointed export packet, README prize mapping, demo video, and social post.
+- Primary goal: finish a countable Build Small submission with real FLUX.2 4B generation, ST3GG scan, optional OpenBMB MiniCPM-V judge evidence, optional NVIDIA Nemotron evidence, checkpointed export packet, README prize mapping, demo video, and social post.
 
 ## Secrets Needed
 
 Do not paste these into chat, commits, logs, or export packets.
 
-- Current Space secret inventory checked after the Raven sprint: `HF_TOKEN` is present; OpenBMB, Nemotron/NVIDIA, and Modal secrets are not present.
-- `HF_TOKEN`: required for gated FLUX.2 Klein 9B access after license acceptance; the app can honestly fall back to the 4B Tiny Titan sidecar if the 9B lane is unavailable.
+- `HF_TOKEN`: optional for public FLUX.2 Klein 4B access and required only if private/gated research lanes are enabled.
 - `MINICPM_BASE_URL`: OpenBMB OpenAI-compatible endpoint base URL.
 - `MINICPM_API_KEY`: OpenBMB bearer token.
 - `MINICPM_MODEL`: default `MiniCPM-V-4.6`.
 - `NEMOTRON_BASE_URL`: OpenAI-compatible Nemotron endpoint if available.
 - `NEMOTRON_API_KEY` or `NVIDIA_API_KEY`: Nemotron provider token.
 - `NEMOTRON_MODEL`: default `nvidia/NVIDIA-Nemotron-Parse-v1.2`.
-- `MODAL_TOKEN_ID` and `MODAL_TOKEN_SECRET`: optional for a documented `netflix/void-model` video repair sidecar job.
 
 ## Verification Commands
 
@@ -35,76 +29,38 @@ python -m compileall app.py src tests
 $env:NEXUS_DISABLE_REAL_HF='1'
 python -c "import app; print('app import ok')"
 $env:PYTEST_DISABLE_PLUGIN_AUTOLOAD='1'
-python -m pytest -q tests -p no:cacheprovider --basetemp=C:\tmp\pytest-nvw-full
+python -m pytest -q tests -p no:cacheprovider
 hf auth whoami --format json
 hf spaces info build-small-hackathon/NEXUS_Visual_Weaver --format json
 ```
 
-If local proxy settings cause HF CLI connection refusal, use a no-proxy Python `urllib` probe with the cached local HF token. Do not print the token.
-
-## HF Space Log Observation
-
-Use the authenticated Hugging Face SSE log APIs for final runtime/build evidence. Do not paste tokens into chat or files; use a local environment variable or the local `huggingface_hub` token cache.
-
-```powershell
-# Container logs
-curl.exe -N -H "Authorization: Bearer $env:HF_TOKEN" `
-  "https://huggingface.co/api/spaces/build-small-hackathon/NEXUS_Visual_Weaver/logs/run"
-
-# Build logs
-curl.exe -N -H "Authorization: Bearer $env:HF_TOKEN" `
-  "https://huggingface.co/api/spaces/build-small-hackathon/NEXUS_Visual_Weaver/logs/build"
-```
-
-If local proxy/TLS settings break PowerShell or `curl`, use a bounded Python/httpx stream with `trust_env=False`, loaded from the local HF token cache. Redact the token from any printed output.
-
-Current evidence from the SSE API:
-
-- `logs/run` returns HTTP 200 and shows the Gradio/MCP startup stream.
-- `logs/build` returns HTTP 200 and shows Build Queued for Space commit `dc6756e`.
-- During the final sprint, HF CLI logs hit local Windows proxy/Unicode issues. Public no-proxy probes verified root HTTP 200 and `/gradio_api/info` callback exposure.
+Avoid pytest `--basetemp=C:\tmp` in this Windows sandbox if `tmp_path` fixtures fail with `PermissionError`. The current tests avoid `tmp_path`.
 
 ## Runtime Flow
 
-1. `run_active_weave` builds the Raven Chronicle run packet from prompt, wardrobe, lore, model stack, and LocateAnything region plan.
-2. FLUX.2 Klein 9B generates the flagship image on Space when HF runtime and gated access are configured; FLUX.2 Klein 4B is an honest sidecar fallback.
+1. `run_active_weave` builds the Raven Chronicle run packet.
+2. FLUX.2 Klein 4B generates the image on Space when HF runtime is enabled.
 3. Generated artifact is scanned by ST3GG.
 4. MiniCPM-V judge runs when OpenBMB secrets are present.
 5. Nemotron evidence runs when Nemotron/NVIDIA endpoint secrets are present.
-6. Modal VOID repair remains a sidecar evidence lane until a real job is documented.
-7. `approve_checkpoint` requires a generated artifact and ST3GG clear/pass state.
-8. `prepare_export_packet` writes a governed JSON packet to `/data/nexus_visual_weaver/exports` or `outputs/exports`.
+6. `approve_checkpoint` requires a generated artifact and ST3GG clear/pass state.
+7. `prepare_export_packet` writes a governed JSON packet to `/data/nexus_visual_weaver/exports` or `outputs/exports`.
 
 ## Claim Rules
 
 - OpenBMB prize claim requires `minicpm_judge.status == "success"` in an export packet.
 - NVIDIA prize claim requires `nemotron_evidence.status == "success"` in an export packet.
 - LocateAnything supports the grounding story but does not replace Nemotron for the NVIDIA prize.
-- Tiny Titan can be claimed only from a successful sidecar export packet because each active sidecar model is <=4B.
-- Raven Quality Stack is the primary story: FLUX.2 Klein 9B, OFFELLIA Q4, LocateAnything, MiniCPM-V, Nemotron, MiniCPM5, FunctionGemma, and Kokoro are individually under 32B.
-- Modal is not claimed unless a real `netflix/void-model` or equivalent Modal job runs and is documented.
+- Tiny Titan can be claimed only from a successful public-demo export packet because each active public model is <=4B.
+- FLUX.2 Klein 9B and OFFELLIA/Gemma remain private research options only.
+- Modal is not claimed unless a real Modal job runs and is documented.
 
 ## Known Risks
 
 - GitHub CLI may fail behind proxy `127.0.0.1:9`; use local git status and HF verification when blocked.
-- Real FLUX generation depends on Space GPU availability and gated 9B access; the 4B sidecar exists to keep the demo useful without mislabeling the flagship lane.
+- Real FLUX generation depends on Space GPU availability and the public 4B runtime loading successfully.
 - OpenBMB and Nemotron endpoints are optional and must show `missing secret` rather than fake success when not configured.
 - Demo video and social post links must be added before final submission.
-- Dev mode served stale `dc6756e` until disabled through the HF API. Keep dev mode off for final judging unless you immediately verify `/config` after re-enabling.
-- The app now supports an explicit ST3GG override reason for evidence-packet export when a generated artifact is review/blocked. This does not mark ST3GG as clear; it records the blocked gate and the human reason.
-
-## Last Verified Checks
-
-- Local compile: `python -m compileall app.py src tests` passed.
-- Local import: `NEXUS_DISABLE_REAL_HF=1 python -c "import app; print('app import ok')"` passed.
-- Local tests: `123 passed, 1 warning`.
-- Secret scan: no `hf_`, `Bearer`, `sk-`, or obvious API-key literals in tracked files.
-- Public Space: root HTTP 200.
-- Public Space API: `/gradio_api/info` exposed `run_active_weave`, `scan_reference`, `approve_checkpoint`, `prepare_export_packet`, and `toggle_adult_catalog`.
-- Public Space config: `Raven Quality Stack` and `FLUX.2 9B PINNED` present; old `Dark Couture v2.4` and `FLUX.2 4B PINNED` absent.
-- Live weave: real FLUX image generated after switching to `Flux2KleinPipeline`; ST3GG marked the generated PNG `review` and blocked export due high entropy review, proving the export gate is active.
-- Readiness: the UI now shows explicit gates for Space/API, Raven Stack, ST3GG, FLUX artifact, checkpoint, export packet, sponsor evidence, and demo/social.
-- Submission assets: see `docs/SUBMISSION_ASSETS.md` for demo script, social draft, thumbnail direction, and final link checklist.
 
 ## Last-Step Checklist
 

src/nexus_visual_weaver/catalog.py

@@ -10,7 +10,7 @@ MODEL_CATALOG: list[ModelCandidate] = [
         role="tiny_titan_sidecar_image_generator",
         task="image-to-image",
         params_b=4.0,
-        runtime="diffusers / public fallback",
+        runtime="diffusers / provider",
         license="apache-2.0",
         source_url="https://hf.co/black-forest-labs/FLUX.2-klein-4B",
     ),
@@ -19,18 +19,19 @@ MODEL_CATALOG: list[ModelCandidate] = [
         role="quality_image_generator",
         task="image-to-image",
         params_b=9.0,
-        runtime="diffusers / gated quality lane",
+        runtime="diffusers / gated provider",
         license="other",
         gated=True,
         source_url="https://hf.co/black-forest-labs/FLUX.2-klein-9B",
     ),
     ModelCandidate(
         repo_id="Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf",
-        role="quality_multimodal_judge",
+        role="private_research_multimodal_judge",
         task="image-text-to-text",
         params_b=12.0,
         runtime="llama.cpp GGUF",
         license="apache-2.0",
+        public_demo=False,
         source_url="https://hf.co/Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf",
     ),
     ModelCandidate(
@@ -80,7 +81,7 @@ MODEL_CATALOG: list[ModelCandidate] = [
     ),
     ModelCandidate(
         repo_id="Brunobkr/OFFELLIA_IQ4_XS_gemma-4-12B-it-heretic",
-        role="adult_private_research_text_judge",
+        role="adult_mode_text_judge",
         task="text-generation",
         params_b=12.0,
         runtime="llama.cpp GGUF",
@@ -88,34 +89,6 @@ MODEL_CATALOG: list[ModelCandidate] = [
         adult_only=True,
         source_url="https://hf.co/Brunobkr/OFFELLIA_IQ4_XS_gemma-4-12B-it-heretic",
     ),
-    ModelCandidate(
-        repo_id="hexgrad/Kokoro-82M",
-        role="audio_lore_tts",
-        task="text-to-speech",
-        params_b=0.082,
-        runtime="local / provider",
-        license="apache-2.0",
-        source_url="https://hf.co/hexgrad/Kokoro-82M",
-    ),
-    ModelCandidate(
-        repo_id="ResembleAI/chatterbox",
-        role="audio_lore_tts_optional",
-        task="text-to-speech",
-        params_b=0.5,
-        runtime="provider / Modal",
-        license="mit",
-        source_url="https://hf.co/ResembleAI/chatterbox",
-    ),
-    ModelCandidate(
-        repo_id="netflix/void-model",
-        role="modal_video_repair",
-        task="video-to-video",
-        params_b=5.0,
-        runtime="Modal / 40GB+ VRAM",
-        license="apache-2.0",
-        public_demo=False,
-        source_url="https://hf.co/netflix/void-model",
-    ),
     ModelCandidate(
         repo_id="Wan-AI/Wan2.2-I2V-A14B-Diffusers",
         role="video_swap_preset",
@@ -139,27 +112,43 @@ MODEL_CATALOG: list[ModelCandidate] = [
 ADAPTER_CATALOG: list[AdapterRecipe] = [
     AdapterRecipe(
         repo_id="DeverStyle/Flux.2-Klein-Loras",
-        adapter_for="black-forest-labs/FLUX.2-klein-4B",
+        adapter_for="black-forest-labs/FLUX.2-klein-9B",
         task="text-to-image style stack",
         license="apache-2.0",
+        trigger_words=["gothic couture", "structured garment"],
+        compatible_repo_ids=[
+            "black-forest-labs/FLUX.2-klein-9B",
+            "black-forest-labs/FLUX.2-klein-4B",
+        ],
     ),
     AdapterRecipe(
         repo_id="fal/flux-2-klein-4B-outpaint-lora",
         adapter_for="black-forest-labs/FLUX.2-klein-4B",
         task="outpaint/inpaint public demo helper",
         license="apache-2.0",
+        compatible_repo_ids=["black-forest-labs/FLUX.2-klein-4B"],
+        requires_image=True,
     ),
     AdapterRecipe(
         repo_id="thedeoxen/refcontrol-FLUX.2-klein-4B-reference-depth-lora",
         adapter_for="black-forest-labs/FLUX.2-klein-base-4B",
         task="reference-depth control for garment layout",
         license="apache-2.0",
+        compatible_repo_ids=[
+            "black-forest-labs/FLUX.2-klein-base-4B",
+            "black-forest-labs/FLUX.2-klein-4B",
+        ],
+        requires_image=True,
     ),
     AdapterRecipe(
         repo_id="nomadoor/flux-2-klein-9B-schematic-lora",
         adapter_for="black-forest-labs/FLUX.2-klein-base-9B",
         task="pose/depth/segmentation schematic control",
         license="other",
+        compatible_repo_ids=[
+            "black-forest-labs/FLUX.2-klein-base-9B",
+            "black-forest-labs/FLUX.2-klein-9B",
+        ],
     ),
     AdapterRecipe(
         repo_id="fal/Qwen-Image-Edit-2511-Multiple-Angles-LoRA",
@@ -195,30 +184,24 @@ ADAPTER_CATALOG: list[AdapterRecipe] = [
     ),
 ]
 
-RAVEN_QUALITY_STACK = [
+DEFAULT_ACTIVE_STACK = [
     "black-forest-labs/FLUX.2-klein-9B",
-    "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf",
     "nvidia/LocateAnything-3B",
     "openbmb/MiniCPM-V-4.6",
     "nvidia/NVIDIA-Nemotron-Parse-v1.2",
     "openbmb/MiniCPM5-1B",
     "onnx-community/functiongemma-270m-it-ONNX",
-    "hexgrad/Kokoro-82M",
 ]
 
-TINY_TITAN_STACK = [
-    "black-forest-labs/FLUX.2-klein-4B",
+PRIVATE_RESEARCH_STACK = [
+    "black-forest-labs/FLUX.2-klein-9B",
+    "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf",
     "nvidia/LocateAnything-3B",
     "openbmb/MiniCPM-V-4.6",
     "nvidia/NVIDIA-Nemotron-Parse-v1.2",
     "openbmb/MiniCPM5-1B",
-    "onnx-community/functiongemma-270m-it-ONNX",
-    "hexgrad/Kokoro-82M",
 ]
 
-DEFAULT_ACTIVE_STACK = RAVEN_QUALITY_STACK
-PRIVATE_RESEARCH_STACK = RAVEN_QUALITY_STACK
-
 
 def filter_catalog(adult_mode: bool = False) -> tuple[list[ModelCandidate], list[AdapterRecipe]]:
     models = [
@@ -233,12 +216,8 @@ def filter_catalog(adult_mode: bool = False) -> tuple[list[ModelCandidate], list
 def active_stack(adult_mode: bool = False) -> list[ModelCandidate]:
     allowed, _ = filter_catalog(adult_mode)
     by_id = {model.repo_id: model for model in allowed}
-    return [by_id[repo_id] for repo_id in RAVEN_QUALITY_STACK if repo_id in by_id]
-
-
-def tiny_titan_stack() -> list[ModelCandidate]:
-    by_id = {model.repo_id: model for model in MODEL_CATALOG}
-    return [by_id[repo_id] for repo_id in TINY_TITAN_STACK if repo_id in by_id]
+    stack_ids = PRIVATE_RESEARCH_STACK if adult_mode else DEFAULT_ACTIVE_STACK
+    return [by_id[repo_id] for repo_id in stack_ids if repo_id in by_id]
 
 
 def parameter_budget(stack: list[ModelCandidate] | None = None) -> dict[str, float | str]:
@@ -259,7 +238,5 @@ def catalog_summary(adult_mode: bool = False) -> dict[str, int | float | str]:
         "models_visible": len(models),
         "adapters_visible": len(adapters),
         "adult_catalog": "enabled" if adult_mode else "hidden",
-        "active_preset": "Raven Quality Stack",
-        "tiny_titan": "sidecar",
         **budget,
     }

src/nexus_visual_weaver/exporter.py

@@ -4,6 +4,7 @@ from __future__ import annotations
 
 import json
 import os
+import re
 import time
 from pathlib import Path
 from typing import Any
@@ -48,6 +49,95 @@ def _artifact_name(output_path: Any) -> str | None:
     return Path(str(output_path)).name
 
 
+SENSITIVE_KEY_RE = re.compile(r"(token|secret|api[_-]?key|authorization|payload_excerpt|raw|base64|bytes)", re.IGNORECASE)
+SECRET_VALUE_RE = re.compile(r"(hf_[A-Za-z0-9]{20,}|Bearer [A-Za-z0-9._-]+|sk-[A-Za-z0-9_-]{20,})")
+CREDENTIAL_NAME_RE = re.compile(r"\b[A-Z0-9_]*(?:API_KEY|TOKEN|SECRET)[A-Z0-9_]*\b")
+WINDOWS_PATH_RE = re.compile(r"[A-Za-z]:[\\/][^\s\"']+")
+
+
+def _sanitize_text(value: str) -> str:
+    text = SECRET_VALUE_RE.sub("[redacted_secret]", value)
+    text = CREDENTIAL_NAME_RE.sub("[redacted_credential_name]", text)
+    text = WINDOWS_PATH_RE.sub(lambda match: f"[local_path]/{Path(match.group(0)).name}", text)
+    repo_text = str(REPO_ROOT)
+    if repo_text in text:
+        text = text.replace(repo_text, "[repo]")
+    if "/data/" in text:
+        text = text.replace("/data/", "[data]/")
+    if len(text) > 1000:
+        text = text[:997] + "..."
+    return text
+
+
+def _safe_dict(value: Any, *, allow_size_bytes: bool = False) -> Any:
+    if isinstance(value, dict):
+        clean: dict[str, Any] = {}
+        for key, item in value.items():
+            key_str = str(key)
+            if SENSITIVE_KEY_RE.search(key_str) and not (allow_size_bytes and key_str == "size_bytes"):
+                continue
+            clean[key_str] = _safe_dict(item, allow_size_bytes=allow_size_bytes)
+        return clean
+    if isinstance(value, list):
+        return [_safe_dict(item, allow_size_bytes=allow_size_bytes) for item in value[:40]]
+    if isinstance(value, str):
+        return _sanitize_text(value)
+    return value
+
+
+def _safe_scan(scan: dict[str, Any]) -> dict[str, Any]:
+    return {
+        "status": scan.get("status"),
+        "scanner": scan.get("scanner"),
+        "export_gate": scan.get("export_gate"),
+        "extension": scan.get("extension"),
+        "magic": scan.get("magic"),
+        "findings": _safe_dict(scan.get("findings") or []),
+        "purification_actions": _safe_dict(scan.get("purification_actions") or []),
+    }
+
+
+def _safe_provider(provider: dict[str, Any] | None) -> dict[str, Any]:
+    provider = provider or {}
+    evidence = provider.get("evidence") if isinstance(provider.get("evidence"), dict) else {}
+    return {
+        "status": provider.get("status"),
+        "provider_state": provider.get("provider_state"),
+        "provider": provider.get("provider"),
+        "repo_id": provider.get("repo_id"),
+        "model": provider.get("model"),
+        "message": _safe_dict(provider.get("message", "")),
+        "evidence": _safe_dict(evidence),
+        "latency_seconds": provider.get("latency_seconds"),
+    }
+
+
+def _safe_reference_metadata(records: Any) -> list[dict[str, Any]]:
+    if not isinstance(records, list):
+        return []
+    cleaned: list[dict[str, Any]] = []
+    for record in records[:20]:
+        if not isinstance(record, dict):
+            continue
+        cleaned.append(
+            {
+                "source": record.get("source"),
+                "status": record.get("status"),
+                "basename": _artifact_name(record.get("basename")) if record.get("basename") else None,
+                "sha256": record.get("sha256"),
+                "size_bytes": record.get("size_bytes"),
+                "st3gg_status": record.get("st3gg_status"),
+                "export_gate": record.get("export_gate"),
+                "magic": record.get("magic"),
+                "extension": record.get("extension"),
+                "domain": record.get("domain"),
+                "url_hash": record.get("url_hash"),
+                "message": _safe_dict(record.get("message", "")),
+            }
+        )
+    return cleaned
+
+
 def export_root() -> Path:
     requested = os.environ.get("NEXUS_EXPORT_DIR")
     candidates = [Path(requested)] if requested else []
@@ -84,49 +174,51 @@ def write_export_packet(
     artifact = _artifact_name(generation.get("output_path"))
     if "output_path" in generation:
         generation["output_path"] = artifact
-    modal_job = operator_state.get("modal_video_repair") or {
-        "status": "deferred",
-        "repo_id": "netflix/void-model",
-        "provider": "modal",
+    generation = _safe_dict(generation)
+    creator_controls = _safe_dict(operator_state.get("creator_controls") or getattr(getattr(run, "request", None), "creator_controls", {}) or {})
+    reference_metadata = _safe_reference_metadata(operator_state.get("reference_metadata") or getattr(getattr(run, "request", None), "reference_metadata", []) or [])
+    st3gg_scan = _safe_scan(scan)
+    minicpm_judge = _safe_provider(operator_state.get("minicpm_judge") or {})
+    nemotron_evidence = _safe_provider(operator_state.get("nemotron_evidence") or {})
+    provider_states = {
+        "generation": generation.get("provider_state"),
+        "minicpm": minicpm_judge.get("status"),
+        "nemotron": nemotron_evidence.get("status"),
+        "operator": operator_state.get("provider_state"),
     }
-    audio_lore = operator_state.get("audio_lore_tts") or {
-        "status": "optional",
-        "repo_id": "hexgrad/Kokoro-82M",
-    }
-    offellia = operator_state.get("offellia_judge") or {
-        "status": "deferred_local",
-        "repo_id": "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf",
-    }
-    tiny_titan = operator_state.get("tiny_titan_sidecar") or {
-        "status": "available",
-        "repo_id": "black-forest-labs/FLUX.2-klein-4B",
-    }
-    locate_grounding = operator_state.get("locateanything_grounding") or {}
-    st3gg_override_reason = str(operator_state.get("st3gg_override_reason", "")).strip()
     packet = {
         "schema": "nexus_visual_weaver.export_packet.v1",
         "run_id": run_id,
         "created_at_epoch": int(time.time()),
-        "active_preset": operator_state.get("active_preset", "Raven Quality Stack"),
         "adult_mode": run_adult_mode,
         "prompt": getattr(getattr(run, "request", None), "prompt", ""),
         "refined_prompt": getattr(getattr(run, "refined_prompt", None), "refined", ""),
         "artifact": artifact,
+        "image_basename": artifact,
+        "creator_controls": creator_controls,
+        "reference_metadata": reference_metadata,
+        "lora_status": {
+            "status": generation.get("lora_status"),
+            "repo_id": generation.get("lora_repo_id"),
+            "message": generation.get("lora_message"),
+        },
         "generation": generation,
-        "st3gg_scan": scan,
-        "st3gg_override_reason": st3gg_override_reason or None,
-        "locateanything_grounding": locate_grounding,
-        "offellia_judge": offellia,
-        "minicpm_judge": operator_state.get("minicpm_judge") or {},
-        "nemotron_evidence": operator_state.get("nemotron_evidence") or {},
-        "modal_video_repair": modal_job,
-        "audio_lore_tts": audio_lore,
-        "tiny_titan_sidecar": tiny_titan,
+        "st3gg_verdict": {
+            "status": st3gg_scan.get("status"),
+            "export_gate": st3gg_scan.get("export_gate"),
+        },
+        "st3gg_scan": st3gg_scan,
+        "minicpm_judge": minicpm_judge,
+        "nemotron_evidence": nemotron_evidence,
         "checkpoint": {
             "status": operator_state.get("checkpoint"),
             "message": operator_state.get("message"),
+            "recommendation": getattr(getattr(run, "checkpoint", None), "recommendation", None),
+            "trust_score": getattr(getattr(run, "checkpoint", None), "trust_score", None),
+            "required_actions": getattr(getattr(run, "checkpoint", None), "required_actions", []),
         },
         "provider_state": operator_state.get("provider_state"),
+        "provider_states": provider_states,
         "model_stack": [
             {
                 "repo_id": model.repo_id,
@@ -142,15 +234,9 @@ def write_export_packet(
             "build_small_32b": budget["status"] == "pass",
             "gradio_space": True,
             "off_brand_custom_ui": True,
-            "openbmb_lane": (operator_state.get("minicpm_judge") or {}).get("status") == "success",
-            "nvidia_nemotron_lane": (operator_state.get("nemotron_evidence") or {}).get("status") == "success",
-            "offellia_quality_lane": offellia.get("status") in {"success", "completed"},
-            "modal_void_lane": modal_job.get("status") in {"success", "completed", "documented"},
-            "tiny_titan_sidecar": tiny_titan.get("status") in {"success", "available", "sidecar"},
-            "raven_quality_stack": True,
-            "locateanything_grounding": bool(locate_grounding.get("targets") or locate_grounding.get("repo_id")),
-            "st3gg_override_recorded": bool(st3gg_override_reason),
-            "st3gg_export_gate": scan.get("export_gate"),
+            "openbmb_lane": minicpm_judge.get("status") == "success",
+            "nvidia_nemotron_lane": nemotron_evidence.get("status") == "success",
+            "st3gg_export_gate": st3gg_scan.get("export_gate"),
         },
     }
     target = export_root() / f"{run_id}.json"

src/nexus_visual_weaver/hf_runtime.py

@@ -3,15 +3,22 @@
 from __future__ import annotations
 
 import os
+import threading
 import time
 from dataclasses import dataclass, asdict
 from pathlib import Path
 from typing import Any
 
+from .catalog import ADAPTER_CATALOG
+from .lora_adapter import load_and_apply, unload_all
+from .schema import AdapterRecipe
+
 
 FLUX_REPO_ID = "black-forest-labs/FLUX.2-klein-9B"
 TINY_TITAN_FLUX_REPO_ID = "black-forest-labs/FLUX.2-klein-4B"
 PRIVATE_RESEARCH_FLUX_REPO_ID = FLUX_REPO_ID
+_PIPELINE_CACHE: dict[str, Any] = {}
+_PIPELINE_CACHE_LOCK = threading.Lock()
 
 
 @dataclass(frozen=True)
@@ -26,6 +33,11 @@ class HFGenerationResult:
     height: int = 1024
     steps: int = 4
     hf_token_present: bool = False
+    lora_status: str = "disabled"
+    lora_repo_id: str | None = None
+    lora_message: str = "No LoRA adapter selected for this run."
+    fallback_used: bool = False
+    primary_error: str | None = None
 
     def to_dict(self) -> dict[str, Any]:
         return asdict(self)
@@ -75,8 +87,51 @@ def active_flux_repo_id() -> str:
     return FLUX_REPO_ID
 
 
-def generate_flux_image(prompt: str, *, seed: int = 0, width: int = 1024, height: int = 1024, steps: int = 4) -> HFGenerationResult:
+def _repo_candidates(repo_id: str) -> list[str]:
+    candidates = [repo_id]
+    if repo_id != TINY_TITAN_FLUX_REPO_ID and os.environ.get("NEXUS_DISABLE_TINY_TITAN_FALLBACK") != "1":
+        candidates.append(TINY_TITAN_FLUX_REPO_ID)
+    return candidates
+
+
+def _get_flux_pipe(repo_id: str, torch_module: Any, pipeline_cls: Any, token: str | None) -> Any:
+    with _PIPELINE_CACHE_LOCK:
+        cached = _PIPELINE_CACHE.get(repo_id)
+        if cached is not None:
+            return cached
+        pipe = pipeline_cls.from_pretrained(repo_id, torch_dtype=torch_module.bfloat16, token=token)
+        pipe.enable_model_cpu_offload()
+        _PIPELINE_CACHE[repo_id] = pipe
+        return pipe
+
+
+def _adapter_recipe(repo_id: str | None) -> AdapterRecipe | None:
+    if not repo_id:
+        return None
+    return next((recipe for recipe in ADAPTER_CATALOG if recipe.repo_id == repo_id), None)
+
+
+def default_lora_repo_id(target_repo_id: str) -> str | None:
+    for recipe in ADAPTER_CATALOG:
+        compatible_ids = {recipe.adapter_for, *recipe.compatible_repo_ids}
+        if recipe.runtime_enabled and not recipe.adult_only and not recipe.requires_image and target_repo_id in compatible_ids:
+            return recipe.repo_id
+    return None
+
+
+def generate_flux_image(
+    prompt: str,
+    *,
+    seed: int = 0,
+    width: int = 1024,
+    height: int = 1024,
+    steps: int = 4,
+    lora_repo_id: str | None = None,
+    adult_mode: bool = False,
+) -> HFGenerationResult:
     repo_id = active_flux_repo_id()
+    selected_lora = lora_repo_id if lora_repo_id is not None else default_lora_repo_id(repo_id)
+    recipe = _adapter_recipe(selected_lora)
     if not hf_runtime_enabled():
         return HFGenerationResult(
             status="disabled",
@@ -87,6 +142,9 @@ def generate_flux_image(prompt: str, *, seed: int = 0, width: int = 1024, height
             height=height,
             steps=steps,
             hf_token_present=bool(_hf_token()),
+            lora_status="disabled",
+            lora_repo_id=recipe.repo_id if recipe else None,
+            lora_message="LoRA loading requires the HF Space GPU runtime.",
         )
 
     started = time.perf_counter()
@@ -103,6 +161,9 @@ def generate_flux_image(prompt: str, *, seed: int = 0, width: int = 1024, height
             height=height,
             steps=steps,
             hf_token_present=bool(_hf_token()),
+            lora_status="disabled",
+            lora_repo_id=recipe.repo_id if recipe else None,
+            lora_message="FLUX runtime import failed before LoRA loading.",
         )
 
     if not torch.cuda.is_available():
@@ -115,49 +176,56 @@ def generate_flux_image(prompt: str, *, seed: int = 0, width: int = 1024, height
             height=height,
             steps=steps,
             hf_token_present=bool(_hf_token()),
+            lora_status="disabled",
+            lora_repo_id=recipe.repo_id if recipe else None,
+            lora_message="CUDA unavailable before LoRA loading.",
         )
 
     token = _hf_token()
-    repo_candidates = [repo_id]
-    if repo_id != TINY_TITAN_FLUX_REPO_ID and os.environ.get("NEXUS_DISABLE_TINY_TITAN_FALLBACK") != "1":
-        repo_candidates.append(TINY_TITAN_FLUX_REPO_ID)
     errors: list[str] = []
-    for candidate in repo_candidates:
+    for candidate in _repo_candidates(repo_id):
         try:
-            dtype = torch.bfloat16
-            pipe = Flux2KleinPipeline.from_pretrained(candidate, torch_dtype=dtype, token=token)
-            pipe.enable_model_cpu_offload()
-            generator = torch.Generator(device="cuda").manual_seed(seed)
-            image = pipe(
-                prompt=prompt,
-                height=height,
-                width=width,
-                guidance_scale=1.0,
-                num_inference_steps=steps,
-                generator=generator,
-            ).images[0]
+            pipe = _get_flux_pipe(candidate, torch, Flux2KleinPipeline, token)
+            if hasattr(pipe, "set_progress_bar_config"):
+                pipe.set_progress_bar_config(disable=True)
+            lora_result = load_and_apply(pipe, recipe, candidate, adult_mode=adult_mode)
+
+            try:
+                generator = torch.Generator(device="cuda").manual_seed(seed)
+                image = pipe(
+                    prompt=prompt,
+                    height=height,
+                    width=width,
+                    guidance_scale=1.0,
+                    num_inference_steps=steps,
+                    generator=generator,
+                ).images[0]
+            finally:
+                unload_all(pipe)
             output_path = _output_dir() / f"nexus_flux_{int(time.time())}_{seed}.png"
             image.save(output_path)
-            fallback = candidate != repo_id
-            message = (
-                f"{candidate} generated a Tiny Titan sidecar artifact after the 9B lane was unavailable."
-                if fallback
-                else f"{candidate} generated a real Raven Quality artifact on HF Space."
-            )
             return HFGenerationResult(
                 status="success",
                 provider_state="generated",
                 repo_id=candidate,
                 output_path=str(output_path),
-                message=message,
+                message=f"{candidate} generated a real Raven Quality artifact on HF Space.",
                 latency_seconds=round(time.perf_counter() - started, 2),
                 width=width,
                 height=height,
                 steps=steps,
                 hf_token_present=bool(token),
+                lora_status=str(lora_result.get("status", "disabled")),
+                lora_repo_id=lora_result.get("repo_id"),
+                lora_message=str(lora_result.get("message", "")),
+                fallback_used=candidate != repo_id,
+                primary_error=errors[0] if candidate != repo_id and errors else None,
             )
         except Exception as exc:  # pragma: no cover - exercised on HF Space with gated/runtime conditions.
             errors.append(f"{candidate}: {_short_error(exc)}")
+            with _PIPELINE_CACHE_LOCK:
+                _PIPELINE_CACHE.pop(candidate, None)
+            continue
     return HFGenerationResult(
         status="error",
         provider_state="blocked",
@@ -167,5 +235,8 @@ def generate_flux_image(prompt: str, *, seed: int = 0, width: int = 1024, height
         width=width,
         height=height,
         steps=steps,
-        hf_token_present=bool(token),
+        hf_token_present=bool(_hf_token()),
+        lora_status="disabled" if recipe is None else "failed",
+        lora_repo_id=recipe.repo_id if recipe else None,
+        lora_message="Generation failed before a usable LoRA evidence state could be produced.",
     )

src/nexus_visual_weaver/lora_adapter.py

@@ -0,0 +1,85 @@
+"""Guarded LoRA adapter loading for HF runtime execution."""
+
+from __future__ import annotations
+
+from dataclasses import asdict
+from typing import Any
+
+from .schema import AdapterRecipe
+
+
+def _status(status: str, recipe: AdapterRecipe | None = None, **extra: Any) -> dict[str, Any]:
+    payload: dict[str, Any] = {
+        "status": status,
+        "repo_id": recipe.repo_id if recipe else None,
+        "adapter_for": recipe.adapter_for if recipe else None,
+        "weight": recipe.weight if recipe else None,
+    }
+    payload.update(extra)
+    return payload
+
+
+def _short_error(exc: BaseException) -> str:
+    text = str(exc).replace("\n", " ").strip()
+    if len(text) > 240:
+        text = text[:237] + "..."
+    return f"{exc.__class__.__name__}: {text}"
+
+
+def adapter_to_dict(recipe: AdapterRecipe) -> dict[str, Any]:
+    return asdict(recipe)
+
+
+def is_compatible(pipe: Any, recipe: AdapterRecipe, target_repo_id: str, *, adult_mode: bool = False) -> bool:
+    if not recipe.runtime_enabled:
+        return False
+    if recipe.adult_only and not adult_mode:
+        return False
+    if recipe.requires_image:
+        return False
+    if not hasattr(pipe, "load_lora_weights"):
+        return False
+
+    compatible_ids = {recipe.adapter_for, *recipe.compatible_repo_ids}
+    if compatible_ids and target_repo_id not in compatible_ids:
+        return False
+    return True
+
+
+def load_and_apply(
+    pipe: Any,
+    recipe: AdapterRecipe | None,
+    target_repo_id: str,
+    *,
+    adult_mode: bool = False,
+    adapter_name: str = "nexus_style",
+) -> dict[str, Any]:
+    if recipe is None:
+        return _status("disabled", message="No LoRA adapter selected for this run.")
+    if recipe.adult_only and not adult_mode:
+        return _status("skipped_incompatible", recipe, message="Adult-only adapter is not available while Adult Mode is off.")
+    if recipe.requires_image:
+        return _status("skipped_incompatible", recipe, message="Adapter requires image-conditioning support that is deferred in P0.")
+    if not hasattr(pipe, "load_lora_weights"):
+        return _status("unsupported_pipeline", recipe, message="Pipeline does not expose load_lora_weights.")
+    if not is_compatible(pipe, recipe, target_repo_id, adult_mode=adult_mode):
+        return _status("skipped_incompatible", recipe, message=f"Adapter is not declared compatible with {target_repo_id}.")
+
+    try:
+        kwargs: dict[str, Any] = {"adapter_name": adapter_name}
+        if recipe.weight_name:
+            kwargs["weight_name"] = recipe.weight_name
+        pipe.load_lora_weights(recipe.repo_id, **kwargs)
+        if hasattr(pipe, "set_adapters"):
+            pipe.set_adapters([adapter_name], adapter_weights=[recipe.weight])
+        return _status("loaded", recipe, message="Adapter loaded and applied for this generation.", adapter_name=adapter_name)
+    except Exception as exc:
+        return _status("failed", recipe, message=_short_error(exc), adapter_name=adapter_name)
+
+
+def unload_all(pipe: Any) -> None:
+    try:
+        if hasattr(pipe, "unload_lora_weights"):
+            pipe.unload_lora_weights()
+    except Exception:
+        return

src/nexus_visual_weaver/model_relay.py

@@ -15,7 +15,6 @@ from typing import Any
 
 PINNED_LANES = {"image_generation", "grounding", "security"}
 ROTATABLE_LANES = {
-    "tiny_titan_sidecar",
     "private_image_research",
     "prompt_router",
     "taste_judge",
@@ -189,11 +188,12 @@ class WeaverModelRelay:
         lane_records = [record for record in self.records.values() if record.lane == lane]
         if lane in PINNED_LANES:
             primary = next((record for record in lane_records if record.pinned), None)
+            fallbacks = [self.records[model_id] for model_id in (primary.fallback_chain if primary else ()) if model_id in self.records]
             return LaneDecision(
                 lane=lane,
                 strategy="pinned",
                 primary=primary,
-                fallbacks=[],
+                fallbacks=fallbacks,
                 reason="Pinned core lane; rotation disabled for creative identity, grounding, or security.",
                 expected_cost_hint=primary.cost_hint if primary else "unavailable",
                 quota_impact=self._quota_impact(primary, now) if primary else {},
@@ -381,7 +381,7 @@ class WeaverModelRelay:
 
     def _decision_reason(self, primary: ModelRecord, strategy: str, public_demo: bool) -> str:
         if strategy == "license_safe_public":
-            return f"{primary.model_id} selected because it is license-safe and within helper budget."
+            return f"{primary.model_id} selected because it is public-demo safe and within helper budget."
         if strategy == "quota_saver":
             return f"{primary.model_id} selected to preserve provider quota and reuse cheaper metadata paths."
         if strategy == "latency_first":
@@ -438,26 +438,26 @@ def default_model_records() -> list[ModelRecord]:
             repo_id="black-forest-labs/FLUX.2-klein-9B",
             license_gate="review_required",
             params_b=9.0,
-            cost_hint="gated_provider_or_quality_space",
+            cost_hint="gated_provider_or_private_space",
             rpm_limit=6,
             rpd_limit=40,
-            quality_score=0.97,
+            quality_score=0.96,
             latency_ms=26000,
             pinned=True,
+            fallback_chain=("flux2-klein-4b-sidecar",),
         ),
         ModelRecord(
-            model_id="flux2-klein-4b-tiny-sidecar",
-            lane="tiny_titan_sidecar",
+            model_id="flux2-klein-4b-sidecar",
+            lane="image_generation",
             provider="hf",
             repo_id="black-forest-labs/FLUX.2-klein-4B",
             license_gate="apache-2.0",
             params_b=4.0,
-            cost_hint="public_fallback_or_tiny_titan_export",
+            cost_hint="provider_or_local",
             rpm_limit=8,
             rpd_limit=60,
             quality_score=0.92,
             latency_ms=21000,
-            fallback_chain=("flux2-klein-9b-quality",),
         ),
         ModelRecord(
             model_id="flux2-klein-9b-private",
@@ -466,12 +466,24 @@ def default_model_records() -> list[ModelRecord]:
             repo_id="black-forest-labs/FLUX.2-klein-9B",
             license_gate="review_required",
             params_b=9.0,
-            cost_hint="legacy_quality_alias",
+            cost_hint="gated_provider_or_private_space",
             rpm_limit=6,
             rpd_limit=40,
             quality_score=0.96,
             latency_ms=26000,
-            health="excluded",
+        ),
+        ModelRecord(
+            model_id="offellia-gemma4-12b-private-image-judge",
+            lane="private_image_research",
+            provider="local",
+            repo_id="Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf",
+            license_gate="private_research",
+            params_b=12.0,
+            cost_hint="local_gpu",
+            rpm_limit=20,
+            rpd_limit=200,
+            quality_score=0.95,
+            latency_ms=4200,
         ),
         ModelRecord(
             model_id="locateanything-3b-anchor",
@@ -774,13 +786,13 @@ def default_model_records() -> list[ModelRecord]:
             model_id="netflix-void-modal",
             lane="video_repair",
             provider="modal",
-            repo_id="netflix/void-model",
-            license_gate="apache-2.0",
-            params_b=5.0,
-            cost_hint="modal_credits_40gb_vram",
-            rpm_limit=4,
-            rpd_limit=30,
-            quality_score=0.88,
+            repo_id="Netflix/VOID",
+            license_gate="private_research",
+            params_b=1.3,
+            cost_hint="modal_credits",
+            rpm_limit=10,
+            rpd_limit=120,
+            quality_score=0.84,
             latency_ms=12000,
             fallback_chain=("void-q5-offline",),
         ),
@@ -788,9 +800,9 @@ def default_model_records() -> list[ModelRecord]:
             model_id="void-q5-offline",
             lane="video_repair",
             provider="local",
-            repo_id="local/netflix-void-q5-video-repair",
+            repo_id="local/VOID-Q5-video-repair",
             license_gate="private_research",
-            params_b=5.0,
+            params_b=1.3,
             cost_hint="offline",
             rpm_limit=20,
             rpd_limit=200,

src/nexus_visual_weaver/planner.py

@@ -17,10 +17,20 @@ def build_command_center_run(
     mode: str = "Strict",
     video_preset: str = "Wan2.2 I2V",
     adult_mode: bool = False,
+    creator_controls: dict | None = None,
+    reference_metadata: list[dict] | None = None,
 ) -> GenerationRun:
-    request = CreativeRequest(prompt=prompt, adult_mode=adult_mode)
+    controls = creator_controls or {}
+    references = reference_metadata or []
+    request = CreativeRequest(
+        prompt=prompt,
+        adult_mode=adult_mode,
+        references=[str(item.get("id") or item.get("basename") or item.get("url_hash")) for item in references],
+        creator_controls=controls,
+        reference_metadata=references,
+    )
     refined = refine_prompt(prompt, adult_mode=adult_mode)
-    outfit = build_outfit_graph(refined.refined, adult_mode=adult_mode)
+    outfit = build_outfit_graph(refined.refined, adult_mode=adult_mode, controls=controls.get("wardrobe") if isinstance(controls, dict) else None)
     inspection = inspect_outfit(outfit)
     lore = build_lore_beats(refined.refined)
     video = build_video_plan(video_preset)
@@ -52,4 +62,3 @@ def build_command_center_run(
         video=video,
         checkpoint=checkpoint,
     )
-

src/nexus_visual_weaver/render.py

@@ -113,13 +113,13 @@ def render_command_header() -> str:
       <div>
         <small>COMMAND INPUT</small>
         <strong>Raven Chronicle Active Weave</strong>
-        <span>Quality stack, reference grounding, model evidence, and checkpoint controls stay in one sticky operator strip.</span>
+        <span>Prompt, reference scan, model route, and checkpoint controls stay in one sticky operator strip.</span>
       </div>
       <div class="nw-command-pills">
-        {badge("RAVEN QUALITY STACK", "pass")}
+        {badge("SFW DEFAULT", "pass")}
         {badge("ST3GG ALWAYS ON", "cyan")}
         {badge("FLUX.2 9B PINNED", "accent")}
-        {badge("4B TINY SIDECAR", "muted")}
+        {badge("4B SIDECAR", "cyan")}
         {badge("HUMAN CHECKPOINT", "warn")}
       </div>
     </section>
@@ -154,56 +154,6 @@ def render_trust_strip(scan: dict | None = None, operator_state: dict | None = N
     """
 
 
-def _submission_link_ready() -> bool:
-    return bool(os.environ.get("NEXUS_DEMO_VIDEO_URL") and os.environ.get("NEXUS_SOCIAL_POST_URL"))
-
-
-def _submission_readiness(scan: dict | None = None, operator_state: dict | None = None) -> tuple[int, int, list[tuple[str, bool, str]]]:
-    scan = scan or {"status": "idle", "export_gate": "pending"}
-    operator_state = operator_state or {}
-    generation = operator_state.get("generation") or {}
-    minicpm = operator_state.get("minicpm_judge") or {}
-    nemotron = operator_state.get("nemotron_evidence") or {}
-    gates = [
-        ("Space/API", True, "Space is serving the command center and MCP/Gradio API."),
-        ("Raven Stack", True, "28.50B quality stack is configured under the 32B rule."),
-        ("ST3GG Visible", True, "Trust strip and export gate are above the fold."),
-        ("FLUX Artifact", generation.get("status") == "success", "Run Active Weave must generate a real artifact."),
-        ("Checkpoint", operator_state.get("checkpoint") == "approved", "Human checkpoint must approve the generated artifact."),
-        ("Export Packet", bool(operator_state.get("export_packet")), "Prepare Export Packet must write evidence JSON."),
-        (
-            "Sponsor Evidence",
-            minicpm.get("status") == "success" and nemotron.get("status") == "success",
-            "MiniCPM and Nemotron both need configured secrets and successful calls.",
-        ),
-        ("Demo/Social", _submission_link_ready(), "Final demo video and social post URLs are still required."),
-    ]
-    complete = sum(1 for _, ok, _ in gates if ok)
-    return complete, len(gates), gates
-
-
-def render_submission_readiness(scan: dict | None = None, operator_state: dict | None = None) -> str:
-    complete, total, gates = _submission_readiness(scan, operator_state)
-    pct = int((complete / total) * 100)
-    chips = "".join(
-        f'<span class="nw-ready-chip {"is-done" if ok else "is-waiting"}">{escape(label)}</span>'
-        for label, ok, _ in gates
-    )
-    blockers = [help_text for _, ok, help_text in gates if not ok][:3]
-    blockers_html = " ".join(f"<span>{escape(text)}</span>" for text in blockers)
-    return f"""
-    <section class="nw-readiness">
-      <div class="nw-ready-head">
-        <strong>Hackathon Readiness</strong>
-        <small>{complete}/{total} gates complete</small>
-      </div>
-      <div class="nw-ready-meter"><i style="width:{pct}%"></i></div>
-      <div class="nw-ready-chips">{chips}</div>
-      <div class="nw-ready-blockers">{blockers_html}</div>
-    </section>
-    """
-
-
 def render_topbar(
     adult_mode: bool = False,
     relay_status: dict | None = None,
@@ -232,7 +182,7 @@ def render_topbar(
     <div class="nw-topbar">
       <div class="nw-brand"><span>NEXUS</span><strong>Visual Weaver</strong></div>
       <div class="nw-topitem"><small>Project</small><strong>Raven Chronicle</strong><i></i></div>
-      <div class="nw-topitem"><small>Active Preset</small><strong>Raven Quality Stack</strong><i></i></div>
+      <div class="nw-topitem"><small>Active Preset</small><strong>Dark Couture v2.4</strong><i></i></div>
       <div class="nw-budget">
         <div><strong>32B Parameter Budget</strong><small>{active:.2f}B / 32B ({pct}%)</small></div>
         <div class="nw-meter"><i style="width:{pct}%"></i></div>
@@ -247,7 +197,6 @@ def render_topbar(
       <div class="nw-locked"><b>18+</b><span>Locked. Enable in Security with explicit justification.</span></div>
     </div>
     {render_trust_strip(scan, operator_state)}
-    {render_submission_readiness(scan, operator_state)}
     """
 
 
@@ -308,7 +257,7 @@ def render_workflow(run: GenerationRun | None = None, operator_state: dict | Non
         "refine": (275, 52, 185, 160, "Refine", ["Prompt Refiner", "Style Harmonizer", "Negative Purge"], "Qwen2.5-7B", "complete", "violet"),
         "judge": (540, 52, 185, 160, "Judge", ["Aesthetic Scorer", "ST3GG Policy Filter", f"Score {score:.2f}"], "MiniCPM / Nemotron", "complete", "blue"),
         "locate": (785, 52, 185, 160, "Locate", ["Reference Locator", "Pose & Composition", "IP-Adapter"], "Refs 3/5", "complete", "cyan"),
-        "generate": (275, 280, 235, 210, "Generate", ["Image / Video Generation", "FLUX.2 9B quality lane", "High-detail couture"], "Steps 4  CFG 1.0", "ready", "green"),
+        "generate": (275, 280, 235, 210, "Generate", ["Image / Video Generation", "FLUX.2 9B + adapter stack", "4B sidecar fallback"], "Steps 4  CFG 1.0", "ready", "green"),
         "video": (590, 280, 235, 210, "Video Path", ["Image to Video", "Frame interpolation", run.video.preset if run else "Wan2.2 / LTX swap"], "Duration 5.6s  24fps", "ready", "blue"),
         "checkpoint": (880, 285, 185, 185, "Human Checkpoint", ["Human review required", "Verify intent, vibe,", "and output before final."], "Review Now", "paused", "amber"),
     }
@@ -621,7 +570,7 @@ def _render_relay_panel(relay_status: dict | None = None) -> str:
     <h3>GMR ModelRelay</h3>
     <ul class="nw-relay">{rows}</ul>
     <div class="nw-relay-foot">
-      {badge("FLUX.2 9B pinned", "pass")} {badge("4B sidecar", "muted")} {badge("LocateAnything pinned", "pass")} {badge(f"dedup hits {dedup_hits}", "muted")}
+      {badge("FLUX.2 9B pinned", "pass")} {badge("4B sidecar", "cyan")} {badge("LocateAnything pinned", "pass")} {badge(f"dedup hits {dedup_hits}", "muted")}
     </div>
     """
 
@@ -642,7 +591,6 @@ def render_provider_cards(relay_status: dict | None = None, adult_mode: bool = F
     optional_statuses = {
         "openbmb": "configured" if _provider_configured("MINICPM_BASE_URL", "MINICPM_API_KEY", "OPENBMB_API_KEY") else "missing secret",
         "nvidia": "configured" if _provider_configured("NEMOTRON_BASE_URL", "NEMOTRON_API_KEY", "NVIDIA_API_KEY") else "missing secret",
-        "modal": "configured" if _env_configured("MODAL_TOKEN_ID", "MODAL_TOKEN_SECRET") else "deferred",
         "fal": "configured" if _env_configured("FAL_KEY") else "blocked",
         "netlify": "configured" if _env_configured("NETLIFY_AUTH_TOKEN", "NETLIFY_SITE_ID", "OPENAI_BASE_URL") else "blocked",
         "cloudflare": "configured" if _env_configured("CLOUDFLARE_API_TOKEN", "CF_ACCOUNT_ID") else "blocked",
@@ -679,9 +627,9 @@ def render_provider_cards(relay_status: dict | None = None, adult_mode: bool = F
             <div class="nw-provider-card nw-provider-optional">
               <small>optional gateway</small>
               <strong>{escape(provider.title())}</strong>
-              <span>{"VOID repair job / Modal credit lane" if provider == "modal" else "off by default / secrets required"}</span>
-              <i class="nw-provider-meter" style="--health:{'74' if state == 'configured' else '42' if state == 'deferred' else '18'}"></i>
-              <div>{badge(state.upper(), "pass" if state == "configured" else "muted" if state == "deferred" else "warn")}{badge("SPONSOR LANE" if provider in {"openbmb", "nvidia", "hf_nvidia", "modal"} else "NOT MVP DEFAULT", "muted")}</div>
+              <span>off by default / secrets required</span>
+              <i class="nw-provider-meter" style="--health:{'74' if state == 'configured' else '18'}"></i>
+              <div>{badge(state.upper(), "pass" if state == "configured" else "warn")}{badge("SPONSOR LANE" if provider in {"openbmb", "nvidia", "hf_nvidia"} else "NOT MVP DEFAULT", "muted")}</div>
             </div>
             """
         )
@@ -726,7 +674,7 @@ def render_inspector(
         scan_status = (scan or {}).get("status", "pass")
     else:
         checks = [(label, True) for label in ["Patent Leather", "Faux Fur", "Lace / Mesh", "Crimson Hardware", "Platform Boots", "Layered Garments"]]
-        model_rows = "<li><span>active stack</span><strong>FLUX.2 9B / OFFELLIA / LocateAnything</strong></li>"
+        model_rows = "<li><span>active stack</span><strong>FLUX.2 9B / MiniCPM / LocateAnything</strong></li>"
         score = 86
         scan_status = (scan or {}).get("status", "pass")
     checks_html = "".join(f'<li><span>{"✓" if ok else "!"}</span>{escape(label)}</li>' for label, ok in checks)
@@ -735,19 +683,11 @@ def render_inspector(
     operator_state = operator_state or {}
     minicpm = operator_state.get("minicpm_judge") or {}
     nemotron = operator_state.get("nemotron_evidence") or {}
-    offellia = operator_state.get("offellia_judge") or {}
-    modal = operator_state.get("modal_video_repair") or {}
-    tts = operator_state.get("audio_lore_tts") or {}
-    locate = operator_state.get("locateanything_grounding") or {}
     sponsor_rows = "".join(
         f"<li><span>{escape(label)}</span><strong>{escape(str(result.get('status', 'pending')).upper())}</strong><em>{escape(str(result.get('repo_id', repo)))}</em></li>"
         for label, repo, result in [
-            ("OFFELLIA Quality Judge", "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf", offellia),
             ("OpenBMB MiniCPM", "openbmb/MiniCPM-V-4.6", minicpm),
             ("NVIDIA Nemotron", "nvidia/NVIDIA-Nemotron-Parse-v1.2", nemotron),
-            ("LocateAnything Grounding", "nvidia/LocateAnything-3B", locate),
-            ("Modal VOID Repair", "netflix/void-model", modal),
-            ("Audio Lore TTS", "hexgrad/Kokoro-82M", tts),
         ]
     )
     findings = [_redact_scan_text(item) for item in (scan.get("findings") or [])]

src/nexus_visual_weaver/schema.py

@@ -17,6 +17,8 @@ class CreativeRequest:
     output_goal: str = "image_to_video"
     adult_mode: bool = False
     references: list[str] = field(default_factory=list)
+    creator_controls: dict[str, Any] = field(default_factory=dict)
+    reference_metadata: list[dict[str, Any]] = field(default_factory=list)
     created_at: str = field(default_factory=utc_now)
 
 
@@ -84,6 +86,11 @@ class AdapterRecipe:
     license: str = "unknown"
     adult_only: bool = False
     compatibility: str = "compatible"
+    weight_name: str | None = None
+    trigger_words: list[str] = field(default_factory=list)
+    compatible_repo_ids: list[str] = field(default_factory=list)
+    runtime_enabled: bool = True
+    requires_image: bool = False
 
 
 @dataclass(frozen=True)

src/nexus_visual_weaver/styles.py

@@ -165,77 +165,6 @@ footer { display: none !important; }
 .nw-trust-card .nw-badge {
   margin-bottom: 1px;
 }
-.nw-readiness {
-  display: grid;
-  grid-template-columns: 190px minmax(180px, 260px) minmax(420px, 1fr);
-  gap: 10px;
-  align-items: center;
-  padding: 8px;
-  border-bottom: 1px solid #171c22;
-  background:
-    linear-gradient(90deg, rgba(255,54,95,.08), transparent 42%),
-    rgba(5, 7, 10, .98);
-}
-.nw-ready-head {
-  min-width: 0;
-}
-.nw-ready-head strong {
-  display: block;
-  font-size: 13px;
-  line-height: 1.2;
-}
-.nw-ready-head small {
-  color: var(--nw-muted);
-  font-size: 11px;
-}
-.nw-ready-meter {
-  height: 10px;
-  border-radius: 999px;
-  overflow: hidden;
-  background: #1d242c;
-  box-shadow: inset 0 0 0 1px rgba(255,255,255,.06);
-}
-.nw-ready-meter i {
-  display: block;
-  height: 100%;
-  background: linear-gradient(90deg, var(--nw-green), var(--nw-cyan));
-  box-shadow: 0 0 16px rgba(32,217,232,.32);
-}
-.nw-ready-chips {
-  display: flex;
-  flex-wrap: wrap;
-  gap: 5px;
-}
-.nw-ready-chip {
-  border: 1px solid var(--nw-line);
-  border-radius: 999px;
-  padding: 4px 8px;
-  color: var(--nw-muted);
-  font-size: 10px;
-  font-weight: 750;
-  white-space: nowrap;
-}
-.nw-ready-chip.is-done {
-  color: #9ff4c6;
-  border-color: rgba(38,215,130,.42);
-  background: rgba(38,215,130,.08);
-}
-.nw-ready-chip.is-waiting {
-  color: #ffb7c4;
-  border-color: rgba(255,54,95,.34);
-  background: rgba(255,54,95,.06);
-}
-.nw-ready-blockers {
-  grid-column: 1 / -1;
-  display: flex;
-  flex-wrap: wrap;
-  gap: 6px;
-}
-.nw-ready-blockers span {
-  color: var(--nw-muted);
-  font-size: 11px;
-  line-height: 1.35;
-}
 .nw-icon { width: 18px; height: 18px; stroke: currentColor; fill: none; stroke-width: 1.8; stroke-linecap: round; stroke-linejoin: round; }
 .nw-shell {
   display: grid;
@@ -1009,8 +938,6 @@ footer { display: none !important; }
 @media (max-width: 1100px) {
   .nw-topbar { grid-template-columns: 1fr; }
   .nw-trust-strip { grid-template-columns: 1fr; }
-  .nw-readiness { grid-template-columns: 1fr; }
-  .nw-ready-blockers { grid-column: 1; }
   .nw-shell { grid-template-columns: 1fr; grid-template-rows: auto; }
   .nw-rail, .nw-inspector, .nw-bottom { grid-column: 1; grid-row: auto; }
   .nw-rail { flex-direction: row; overflow-x: auto; }

src/nexus_visual_weaver/wardrobe.py

@@ -17,10 +17,23 @@ SLOT_BLUEPRINTS: list[tuple[str, str, str, str, str]] = [
 ]
 
 
-def build_outfit_graph(prompt: str, adult_mode: bool = False) -> OutfitGraph:
+def build_outfit_graph(prompt: str, adult_mode: bool = False, controls: dict | None = None) -> OutfitGraph:
     lowered = prompt.lower()
+    controls = controls or {}
+    locked_slots = set(controls.get("locked_slots") or [])
+    locate_focus = set(controls.get("locate_focus") or [])
     slots: list[WardrobeSlot] = []
     for index, (name, description, material, palette, lora_hint) in enumerate(SLOT_BLUEPRINTS):
+        if name == "outerwear" and controls.get("outerwear"):
+            description = str(controls["outerwear"])
+        if name == "upper_body" and controls.get("upper_body"):
+            material = str(controls["upper_body"])
+        if name == "footwear" and controls.get("footwear"):
+            description = str(controls["footwear"])
+        if name == "jewelry" and controls.get("hardware"):
+            material = str(controls["hardware"])
+        if controls.get("palette"):
+            palette = str(controls["palette"])
         locked = any(token in lowered for token in name.split("_")) or material.replace("_", " ") in lowered
         if name == "outerwear" and "coat" in lowered:
             locked = True
@@ -28,6 +41,8 @@ def build_outfit_graph(prompt: str, adult_mode: bool = False) -> OutfitGraph:
             locked = True
         if name == "jewelry" and ("crimson" in lowered or "hardware" in lowered):
             locked = True
+        if name in locked_slots:
+            locked = True
         slots.append(
             WardrobeSlot(
                 name=name,
@@ -37,11 +52,10 @@ def build_outfit_graph(prompt: str, adult_mode: bool = False) -> OutfitGraph:
                 lora_hint=lora_hint,
                 locked=locked,
                 adult_only=False if not adult_mode else name in {"upper_body", "lower_body"},
-                locate_region="auto-map",
+                locate_region="manual-focus" if name in locate_focus else "auto-map",
                 edit_priority=max(1, 5 - index // 2),
             )
         )
     locked_count = sum(1 for slot in slots if slot.locked)
     score = round(0.68 + min(0.24, locked_count * 0.035), 2)
     return OutfitGraph(slots=slots, score=score)
-

tests/test_app_callbacks.py

@@ -97,6 +97,36 @@ def test_run_weave_returns_stateful_dashboard_packet() -> None:
     assert result[13].checkpoint.checkpoint_id.startswith("nw-")
     assert result[15]["provider_state"] == "checkpointed"
     assert result[16]["interactive"] is True
+    assert result[15]["creator_controls"]["wardrobe"]["footwear"] == "platform boots"
+    assert result[15]["generation"]["lora_status"] == "disabled"
+
+
+def test_run_weave_persists_additive_creator_controls_and_reference_url_metadata() -> None:
+    result = app.run_weave(
+        "gothic patent leather platform boots, crimson hardware",
+        "Frontier",
+        "LTX-2.3",
+        False,
+        None,
+        "Wardrobe",
+        "layered tactical silhouette",
+        "faux fur collar coat",
+        "black mesh layer",
+        "patent leather heels",
+        "obsidian, pearl, crimson",
+        "silver occult buckles",
+        "https://shop.example.test/item/123",
+    )
+
+    state = result[15]
+    run = result[13]
+
+    assert state["creator_controls"]["reasoning_mode"] == "Frontier"
+    assert state["creator_controls"]["wardrobe"]["footwear"] == "patent leather heels"
+    assert state["reference_metadata"][0]["source"] == "url"
+    assert state["reference_metadata"][0]["domain"] == "shop.example.test"
+    assert "url_hash" in state["reference_metadata"][0]
+    assert run.request.creator_controls["wardrobe"]["outerwear"] == "faux fur collar coat"
 
 
 def test_operator_actions_transition_checkpoint_export_and_stop() -> None:
@@ -146,40 +176,6 @@ def test_export_blocks_without_checkpoint() -> None:
     assert "checkpoint" in blocked[13]["message"].lower()
 
 
-def test_export_allows_st3gg_review_with_explicit_override() -> None:
-    from pathlib import Path
-
-    result = app.run_weave(
-        "gothic patent leather platform boots, crimson hardware",
-        "Strict",
-        "Wan2.2 I2V",
-        False,
-        None,
-        "Forge",
-    )
-    run = result[13]
-    artifact_path = Path("outputs/test-override-artifact.png")
-    artifact_path.parent.mkdir(parents=True, exist_ok=True)
-    artifact_path.write_bytes(b"\x89PNG\r\n\x1a\n\x00\x00\x00\rIEND\xaeB`\x82NEXUS_REVIEW")
-    state = {
-        **result[15],
-        "generation": {**result[15]["generation"], "status": "success", "output_path": str(artifact_path)},
-        "checkpoint": "approved",
-        "provider_state": "checkpointed",
-    }
-    review_scan = {"status": "review", "export_gate": "blocked", "findings": ["high entropy sample"], "purification_actions": ["manual review"]}
-
-    blocked = app.export_packet(run, False, review_scan, "Forge", state, "")
-    assert blocked[13]["provider_state"] == "blocked"
-    assert "override reason" in blocked[13]["message"].lower()
-
-    exported = app.export_packet(run, False, review_scan, "Forge", state, "Judge-reviewed generated PNG; evidence packet only.")
-    assert exported[13]["provider_state"] == "exported"
-    assert exported[13]["export"] == "override"
-    assert exported[13]["st3gg_override_reason"] == "Judge-reviewed generated PNG; evidence packet only."
-    assert "export_packet" in exported[13]
-
-
 def test_reference_scan_cannot_clear_blocked_generated_artifact() -> None:
     base = app.ROOT / "outputs" / "test-app-callbacks"
     base.mkdir(parents=True, exist_ok=True)

tests/test_command_center.py

@@ -3,13 +3,10 @@ from pathlib import Path
 from nexus_visual_weaver.catalog import (
     DEFAULT_ACTIVE_STACK,
     PRIVATE_RESEARCH_STACK,
-    RAVEN_QUALITY_STACK,
-    TINY_TITAN_STACK,
     active_stack,
     catalog_summary,
     filter_catalog,
     parameter_budget,
-    tiny_titan_stack,
 )
 from nexus_visual_weaver.grounding import inspect_outfit
 from nexus_visual_weaver.model_relay import WeaverModelRelay
@@ -20,7 +17,6 @@ from nexus_visual_weaver.render import (
     render_inspector,
     render_operations_panel,
     render_provider_cards,
-    render_submission_readiness,
     render_topbar,
     render_trust_strip,
 )
@@ -67,25 +63,16 @@ def test_active_parameter_budget_passes_default_stack() -> None:
     assert budget["active_b"] <= 32.0
 
 
-def test_default_stack_uses_raven_quality_models() -> None:
+def test_public_stack_uses_flux_9b_and_excludes_offellia() -> None:
     stack = active_stack(False)
     repo_ids = {model.repo_id for model in stack}
 
     assert "black-forest-labs/FLUX.2-klein-9B" in repo_ids
-    assert "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf" in repo_ids
     assert "black-forest-labs/FLUX.2-klein-4B" not in repo_ids
+    assert not any("OFFELLIA" in repo_id for repo_id in repo_ids)
     assert all(model.params_b < 32.0 for model in stack)
 
 
-def test_tiny_titan_stack_is_sidecar_only_and_all_models_are_under_4b() -> None:
-    stack = tiny_titan_stack()
-    repo_ids = {model.repo_id for model in stack}
-
-    assert "black-forest-labs/FLUX.2-klein-4B" in repo_ids
-    assert "black-forest-labs/FLUX.2-klein-9B" not in repo_ids
-    assert all(model.params_b <= 4.0 for model in stack)
-
-
 def test_private_catalog_keeps_stronger_research_models_available() -> None:
     private_models, _ = filter_catalog(True)
     repo_ids = {model.repo_id for model in private_models}
@@ -227,7 +214,7 @@ def test_command_header_exposes_governed_run_controls() -> None:
     assert "Raven Chronicle Active Weave" in header
     assert "ST3GG ALWAYS ON" in header
     assert "FLUX.2 9B PINNED" in header
-    assert "4B TINY SIDECAR" in header
+    assert "4B SIDECAR" in header
     assert "HUMAN CHECKPOINT" in header
 
 
@@ -316,8 +303,6 @@ def test_render_topbar_includes_trust_strip() -> None:
 
     assert "TRUST MODEL" in html
     assert "nw-trust-strip" in html
-    assert "Hackathon Readiness" in html
-    assert "3/8 gates complete" in html
 
 
 def test_render_topbar_with_scan_passes_to_trust_strip() -> None:
@@ -342,24 +327,6 @@ def test_render_topbar_default_scan_shows_fixture_evidence() -> None:
     assert "PNG trailing bytes -> blocked." in html
 
 
-def test_submission_readiness_tracks_run_and_submission_blockers() -> None:
-    html = render_submission_readiness(
-        scan={"status": "review", "export_gate": "blocked"},
-        operator_state={
-            "generation": {"status": "success"},
-            "checkpoint": "approved",
-            "export_packet": {"path": "outputs/exports/nw.json"},
-            "minicpm_judge": {"status": "missing_secret"},
-            "nemotron_evidence": {"status": "missing_secret"},
-        },
-    )
-
-    assert "Hackathon Readiness" in html
-    assert "6/8 gates complete" in html
-    assert "MiniCPM and Nemotron" in html
-    assert "Final demo video and social post URLs" in html
-
-
 # --- render_inspector sponsor evidence tests ---
 
 def test_render_inspector_shows_sponsor_evidence_section() -> None:
@@ -393,7 +360,7 @@ def test_render_inspector_with_success_judge_shows_success_status() -> None:
 def test_render_inspector_shows_default_stack_label_without_run() -> None:
     html = render_inspector()
 
-    assert "FLUX.2 9B / OFFELLIA / LocateAnything" in html
+    assert "FLUX.2 9B / MiniCPM / LocateAnything" in html
 
 
 # --- render_provider_cards sponsor lane tests ---
@@ -480,11 +447,11 @@ def test_filter_catalog_includes_flux_9b_in_adult_mode() -> None:
     assert "black-forest-labs/FLUX.2-klein-9B" in repo_ids
 
 
-def test_filter_catalog_includes_offellia_in_public_mode() -> None:
+def test_filter_catalog_excludes_offellia_in_public_mode() -> None:
     models, _ = filter_catalog(False)
     repo_ids = {model.repo_id for model in models}
 
-    assert "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf" in repo_ids
+    assert not any("OFFELLIA" in repo_id for repo_id in repo_ids)
 
 
 def test_active_stack_uses_private_research_stack_in_adult_mode() -> None:
@@ -492,7 +459,6 @@ def test_active_stack_uses_private_research_stack_in_adult_mode() -> None:
     repo_ids = {model.repo_id for model in stack}
 
     assert "black-forest-labs/FLUX.2-klein-9B" in repo_ids
-    assert "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf" in repo_ids
     assert "black-forest-labs/FLUX.2-klein-4B" not in repo_ids
 
 
@@ -502,13 +468,10 @@ def test_private_research_stack_constant_contains_9b_and_offellia() -> None:
 
 
 def test_default_active_stack_constant_uses_9b_and_sponsor_models() -> None:
-    assert DEFAULT_ACTIVE_STACK == RAVEN_QUALITY_STACK
     assert "black-forest-labs/FLUX.2-klein-9B" in DEFAULT_ACTIVE_STACK
-    assert "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf" in DEFAULT_ACTIVE_STACK
     assert "openbmb/MiniCPM-V-4.6" in DEFAULT_ACTIVE_STACK
     assert "nvidia/NVIDIA-Nemotron-Parse-v1.2" in DEFAULT_ACTIVE_STACK
     assert "black-forest-labs/FLUX.2-klein-4B" not in DEFAULT_ACTIVE_STACK
-    assert "black-forest-labs/FLUX.2-klein-4B" in TINY_TITAN_STACK
 
 
 # --- schema ModelCandidate public_demo tests ---
@@ -548,8 +511,5 @@ def test_public_demo_false_models_are_excluded_from_public_filter() -> None:
         license="other",
         public_demo=False,
     )
-    # public_demo=False should mean filter_catalog(False) excludes hidden support lanes.
-    # The catalog-level test: verify Modal VOID is absent from public catalog scope.
     models_public, _ = filter_catalog(False)
     assert all(m.public_demo for m in models_public)
-    assert "netflix/void-model" not in {model.repo_id for model in models_public}

tests/test_exporter.py

@@ -10,14 +10,31 @@ def _make_base_state(**overrides):
         "provider_state": "export_ready",
         "checkpoint": "approved",
         "message": "approved",
-        "generation": {"status": "success", "output_path": "/data/artifact.png", "hf_token_present": True},
-        "locateanything_grounding": {"status": "pass", "repo_id": "nvidia/LocateAnything-3B", "targets": [{"slot_name": "footwear"}]},
-        "offellia_judge": {"status": "deferred_local", "repo_id": "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf"},
+        "generation": {
+            "status": "success",
+            "provider_state": "generated",
+            "output_path": "/data/artifact.png",
+            "hf_token_present": True,
+            "lora_status": "loaded",
+            "lora_repo_id": "DeverStyle/Flux.2-Klein-Loras",
+            "lora_message": "Adapter loaded and applied for this generation.",
+        },
+        "creator_controls": {
+            "reasoning_mode": "Strict",
+            "wardrobe": {"outerwear": "black patent leather long coat", "footwear": "platform boots"},
+        },
+        "reference_metadata": [
+            {
+                "source": "upload",
+                "basename": "C:/Users/speci.000/Downloads/reference.png",
+                "sha256": "a" * 64,
+                "size_bytes": 128,
+                "st3gg_status": "pass",
+                "export_gate": "clear",
+            }
+        ],
         "minicpm_judge": {"status": "success", "repo_id": "openbmb/MiniCPM-V-4.6"},
         "nemotron_evidence": {"status": "missing_secret", "repo_id": "nvidia/NVIDIA-Nemotron-Parse-v1.2"},
-        "modal_video_repair": {"status": "deferred", "repo_id": "netflix/void-model", "provider": "modal"},
-        "audio_lore_tts": {"status": "optional", "repo_id": "hexgrad/Kokoro-82M"},
-        "tiny_titan_sidecar": {"status": "available", "repo_id": "black-forest-labs/FLUX.2-klein-4B"},
     }
     state.update(overrides)
     return state
@@ -37,17 +54,13 @@ def test_write_export_packet_records_evidence_without_secrets(monkeypatch) -> No
     assert payload["hackathon_claims"]["openbmb_lane"] is True
     assert payload["hackathon_claims"]["nvidia_nemotron_lane"] is False
     assert payload["parameter_budget"]["status"] == "pass"
-    assert payload["active_preset"] == "Raven Quality Stack"
-    assert payload["modal_video_repair"]["repo_id"] == "netflix/void-model"
-    assert payload["offellia_judge"]["repo_id"] == "Brunobkr/OFFELLIA_Q4_0_gemma-4-12B-it.gguf"
-    assert payload["audio_lore_tts"]["repo_id"] == "hexgrad/Kokoro-82M"
-    assert payload["tiny_titan_sidecar"]["repo_id"] == "black-forest-labs/FLUX.2-klein-4B"
-    assert payload["hackathon_claims"]["raven_quality_stack"] is True
-    assert payload["hackathon_claims"]["locateanything_grounding"] is True
-    assert payload["hackathon_claims"]["offellia_quality_lane"] is False
     assert "token" not in json.dumps(payload).lower()
     assert payload["artifact"] == "artifact.png"
+    assert payload["image_basename"] == "artifact.png"
     assert payload["generation"]["output_path"] == "artifact.png"
+    assert payload["lora_status"]["status"] == "loaded"
+    assert payload["creator_controls"]["wardrobe"]["footwear"] == "platform boots"
+    assert payload["reference_metadata"][0]["basename"] == "reference.png"
     assert "/data/" not in json.dumps(payload)
 
 
@@ -131,20 +144,6 @@ def test_export_packet_hackathon_claims_st3gg_export_gate(monkeypatch) -> None:
     assert payload_blocked["hackathon_claims"]["st3gg_export_gate"] == "blocked"
 
 
-def test_export_packet_records_st3gg_override_reason(monkeypatch) -> None:
-    monkeypatch.setenv("NEXUS_EXPORT_DIR", "outputs/test-exports")
-    run = build_command_center_run("review override brief")
-    scan = {"status": "review", "export_gate": "blocked"}
-    state = _make_base_state(st3gg_override_reason="Human reviewed generated artifact for evidence-only export.")
-
-    result = write_export_packet(run=run, scan=scan, operator_state=state, adult_mode=False)
-    payload = json.loads(Path(result["path"]).read_text(encoding="utf-8"))
-
-    assert payload["st3gg_override_reason"] == "Human reviewed generated artifact for evidence-only export."
-    assert payload["hackathon_claims"]["st3gg_override_recorded"] is True
-    assert payload["hackathon_claims"]["st3gg_export_gate"] == "blocked"
-
-
 def test_export_packet_includes_model_stack_and_prompts(monkeypatch) -> None:
     monkeypatch.setenv("NEXUS_EXPORT_DIR", "outputs/test-exports")
     run = build_command_center_run("raven archivist couture brief")
@@ -211,3 +210,62 @@ def test_export_packet_returns_both_path_and_packet(monkeypatch) -> None:
     assert "packet" in result
     assert isinstance(result["packet"], dict)
     assert result["packet"]["schema"] == "nexus_visual_weaver.export_packet.v1"
+
+
+def test_export_packet_redacts_raw_payload_paths_and_secret_like_values(monkeypatch) -> None:
+    monkeypatch.setenv("NEXUS_EXPORT_DIR", "outputs/test-exports")
+    run = build_command_center_run(
+        "redaction brief",
+        creator_controls={"wardrobe": {"footwear": "platform boots"}},
+        reference_metadata=[{"source": "url", "domain": "shop.example.test", "url_hash": "b" * 64}],
+    )
+    scan = {
+        "status": "review",
+        "export_gate": "blocked",
+        "payload_excerpt": "hf_" + "abcdefghijklmnopqrstuvwxyz123456",
+        "findings": ["raw hidden bytes at C:/Users/speci.000/Downloads/thing.png"],
+        "purification_actions": ["base64 payload removed from /data/nexus_visual_weaver/artifact.png"],
+    }
+    state = _make_base_state(
+        generation={
+            "status": "success",
+            "provider_state": "generated",
+            "output_path": "/data/nexus_visual_weaver/artifact.png",
+            "hf_token_present": True,
+            "lora_status": "failed",
+            "lora_repo_id": "example/lora",
+            "lora_message": "RuntimeError: failed without token value",
+        },
+        minicpm_judge={
+            "status": "failed",
+            "provider_state": "failed",
+            "provider": "OpenBMB",
+            "repo_id": "openbmb/MiniCPM-V-4.6",
+            "model": "MiniCPM-V-4.6",
+            "message": "Set MINICPM_API_KEY in Space secrets.",
+            "evidence": {"raw_summary": "Bearer " + "secret-value", "configured": True},
+        },
+        reference_metadata=[
+            {
+                "source": "upload",
+                "basename": "C:/Users/speci.000/Downloads/reference.png",
+                "sha256": "c" * 64,
+                "size_bytes": 2048,
+                "st3gg_status": "review",
+                "export_gate": "blocked",
+            }
+        ],
+    )
+
+    result = write_export_packet(run=run, scan=scan, operator_state=state, adult_mode=False)
+    payload = json.loads(Path(result["path"]).read_text(encoding="utf-8"))
+    serialized = json.dumps(payload)
+
+    assert "payload_excerpt" not in serialized
+    assert "raw_summary" not in serialized
+    assert "Bearer " + "secret-value" not in serialized
+    assert "MINICPM_API_KEY" not in serialized
+    assert "/data/" not in serialized
+    assert "C:/Users/speci.000/Downloads" not in serialized
+    assert payload["st3gg_verdict"] == {"status": "review", "export_gate": "blocked"}
+    assert payload["provider_states"]["minicpm"] == "failed"

tests/test_hf_runtime.py

@@ -4,7 +4,7 @@ from nexus_visual_weaver.hf_runtime import (
     FLUX_REPO_ID,
     PRIVATE_RESEARCH_FLUX_REPO_ID,
     TINY_TITAN_FLUX_REPO_ID,
-    active_flux_repo_id,
+    default_lora_repo_id,
     generate_flux_image,
     hf_runtime_enabled,
 )
@@ -22,19 +22,18 @@ def test_hf_runtime_is_disabled_locally_by_default(monkeypatch) -> None:
     assert result.status == "disabled"
     assert result.provider_state == "dry-run"
     assert result.repo_id == "black-forest-labs/FLUX.2-klein-9B"
+    assert result.lora_status == "disabled"
+    assert result.lora_repo_id == "DeverStyle/Flux.2-Klein-Loras"
 
 
-def test_quality_and_sidecar_flux_repo_ids_are_split(monkeypatch) -> None:
-    monkeypatch.delenv("NEXUS_FLUX_REPO_ID", raising=False)
-    monkeypatch.delenv("NEXUS_TINY_TITAN_MODE", raising=False)
-
+def test_flux_repo_ids_use_9b_with_4b_sidecar() -> None:
     assert FLUX_REPO_ID == "black-forest-labs/FLUX.2-klein-9B"
     assert PRIVATE_RESEARCH_FLUX_REPO_ID == "black-forest-labs/FLUX.2-klein-9B"
     assert TINY_TITAN_FLUX_REPO_ID == "black-forest-labs/FLUX.2-klein-4B"
-    assert active_flux_repo_id() == FLUX_REPO_ID
 
-    monkeypatch.setenv("NEXUS_TINY_TITAN_MODE", "1")
-    assert active_flux_repo_id() == TINY_TITAN_FLUX_REPO_ID
+
+def test_default_lora_repo_id_prefers_runtime_enabled_compatible_adapter() -> None:
+    assert default_lora_repo_id("black-forest-labs/FLUX.2-klein-9B") == "DeverStyle/Flux.2-Klein-Loras"
 
 
 def test_artifact_lane_embeds_generated_image() -> None:

tests/test_lora_adapter.py

@@ -0,0 +1,90 @@
+from nexus_visual_weaver.lora_adapter import load_and_apply, unload_all
+from nexus_visual_weaver.schema import AdapterRecipe
+
+
+class FakeLoraPipe:
+    def __init__(self) -> None:
+        self.loaded: list[tuple[str, dict]] = []
+        self.adapters: tuple[list[str], list[float]] | None = None
+        self.unloaded = False
+
+    def load_lora_weights(self, repo_id: str, **kwargs) -> None:
+        self.loaded.append((repo_id, kwargs))
+
+    def set_adapters(self, adapter_names: list[str], adapter_weights: list[float]) -> None:
+        self.adapters = (adapter_names, adapter_weights)
+
+    def unload_lora_weights(self) -> None:
+        self.unloaded = True
+
+
+class FailingLoraPipe(FakeLoraPipe):
+    def load_lora_weights(self, repo_id: str, **kwargs) -> None:
+        raise RuntimeError("adapter load failed")
+
+
+class UnsupportedPipe:
+    pass
+
+
+def test_load_and_apply_reports_loaded_for_compatible_pipe() -> None:
+    recipe = AdapterRecipe(
+        repo_id="example/style-lora",
+        adapter_for="black-forest-labs/FLUX.2-klein-9B",
+        task="style",
+        weight=0.6,
+    )
+    pipe = FakeLoraPipe()
+
+    result = load_and_apply(pipe, recipe, "black-forest-labs/FLUX.2-klein-9B")
+
+    assert result["status"] == "loaded"
+    assert result["repo_id"] == "example/style-lora"
+    assert pipe.loaded == [("example/style-lora", {"adapter_name": "nexus_style"})]
+    assert pipe.adapters == (["nexus_style"], [0.6])
+
+
+def test_load_and_apply_reports_skipped_for_incompatible_adapter() -> None:
+    recipe = AdapterRecipe(
+        repo_id="example/style-lora",
+        adapter_for="black-forest-labs/FLUX.2-klein-4B",
+        task="style",
+    )
+
+    result = load_and_apply(FakeLoraPipe(), recipe, "black-forest-labs/FLUX.2-klein-9B")
+
+    assert result["status"] == "skipped_incompatible"
+    assert "not declared compatible" in result["message"]
+
+
+def test_load_and_apply_reports_unsupported_pipeline() -> None:
+    recipe = AdapterRecipe(
+        repo_id="example/style-lora",
+        adapter_for="black-forest-labs/FLUX.2-klein-9B",
+        task="style",
+    )
+
+    result = load_and_apply(UnsupportedPipe(), recipe, "black-forest-labs/FLUX.2-klein-9B")
+
+    assert result["status"] == "unsupported_pipeline"
+
+
+def test_load_and_apply_reports_failed_without_raising() -> None:
+    recipe = AdapterRecipe(
+        repo_id="example/style-lora",
+        adapter_for="black-forest-labs/FLUX.2-klein-9B",
+        task="style",
+    )
+
+    result = load_and_apply(FailingLoraPipe(), recipe, "black-forest-labs/FLUX.2-klein-9B")
+
+    assert result["status"] == "failed"
+    assert "RuntimeError" in result["message"]
+
+
+def test_unload_all_uses_pipeline_unload_hook() -> None:
+    pipe = FakeLoraPipe()
+
+    unload_all(pipe)
+
+    assert pipe.unloaded is True

tests/test_model_relay.py

@@ -13,18 +13,17 @@ def test_pinned_lanes_do_not_rotate() -> None:
     assert decision.primary is not None
     assert decision.primary.repo_id == "black-forest-labs/FLUX.2-klein-9B"
     assert decision.primary.params_b == 9.0
-    assert decision.fallbacks == []
+    assert [fallback.repo_id for fallback in decision.fallbacks] == ["black-forest-labs/FLUX.2-klein-4B"]
     assert "rotation disabled" in decision.reason
 
 
-def test_tiny_titan_sidecar_keeps_flux_4b_available() -> None:
+def test_private_image_research_keeps_flux_9b_available() -> None:
     relay = WeaverModelRelay()
-    decision = relay.select_lane("tiny_titan_sidecar", budget=4.0, public_demo=True, strategy="license_safe_public")
+    decision = relay.select_lane("private_image_research", budget=9.0, public_demo=False, strategy="private_research")
 
     assert decision.primary is not None
-    assert decision.primary.repo_id == "black-forest-labs/FLUX.2-klein-4B"
+    assert decision.primary.repo_id == "black-forest-labs/FLUX.2-klein-9B"
     assert decision.primary.pinned is False
-    assert decision.primary.params_b == 4.0
 
 
 def test_public_private_taste_judge_respects_license_and_budget() -> None:
@@ -137,8 +136,6 @@ def test_optional_external_gateways_are_registered_but_excluded_by_default() ->
     assert relay.records["netlify-ai-gateway-helper"].provider == "netlify"
     assert relay.records["cloudflare-agent-helper"].provider == "cloudflare"
     assert relay.records["fal-media-adapter"].provider == "fal"
-    assert relay.records["netflix-void-modal"].repo_id == "netflix/void-model"
-    assert relay.records["netflix-void-modal"].params_b == 5.0
     assert relay.records["netflix-void-modal"].health == "healthy"
     assert relay.records["netlify-ai-gateway-helper"].health == "excluded"
     assert relay.records["fal-media-adapter"].health == "excluded"
@@ -188,27 +185,31 @@ def test_private_image_research_lane_is_rotatable() -> None:
     assert "private_image_research" in ROTATABLE_LANES
 
 
-def test_flux2_klein_9b_is_pinned_quality_lane() -> None:
+def test_flux2_klein_9b_is_pinned_with_4b_sidecar_fallback() -> None:
     relay = WeaverModelRelay()
 
     record = relay.records["flux2-klein-9b-quality"]
+    fallback = relay.records["flux2-klein-4b-sidecar"]
 
     assert record.lane == "image_generation"
     assert record.pinned is True
     assert record.params_b == 9.0
     assert record.license_gate == "review_required"
     assert record.repo_id == "black-forest-labs/FLUX.2-klein-9B"
+    assert "flux2-klein-4b-sidecar" in record.fallback_chain
+    assert fallback.repo_id == "black-forest-labs/FLUX.2-klein-4B"
+    assert fallback.pinned is False
 
 
-def test_flux2_klein_4b_is_tiny_titan_sidecar() -> None:
+def test_flux2_klein_9b_is_not_pinned_and_in_private_research() -> None:
     relay = WeaverModelRelay()
 
-    record = relay.records["flux2-klein-4b-tiny-sidecar"]
+    record = relay.records["flux2-klein-9b-private"]
 
-    assert record.lane == "tiny_titan_sidecar"
+    assert record.lane == "private_image_research"
     assert record.pinned is False
-    assert record.params_b == 4.0
-    assert record.license_gate == "apache-2.0"
+    assert record.params_b == 9.0
+    assert record.license_gate == "review_required"
 
 
 def test_minicpm_has_fallback_chain_configured() -> None: