SceneConductor

One-shot installer that takes a FRESH clone of SceneConductor to a fully runnable pipeline — initializes submodules, downloads Blender 4.2.1, builds all five conda envs (./setup.sh, including the GroundingDINO torch-compat patch + _C verification), downloads every model checkpoint (GroundedSAM, GALP, SAM3D, Qwen), wires the SAM3D symlink, and verifies everything. Use this skill WHENEVER the user wants to install / set up / provision / bootstrap SceneConductor, is starting from a fresh git clone, hits setup errors such as a missing `groundingdino._C` CUDA extension, a conda env that cannot be found, or absent model checkpoints, or asks "how do I get this running". Trigger on "/sceneconductor-setup", "set up SceneConductor", "install the pipeline", "provision the environment", "bootstrap a fresh clone". Prefer this over hand-running setup.sh — it also fetches checkpoints and validates the install, which setup.sh alone does not.

Initialize a 3D scene from a single indoor `image.png` — produces segmentation masks, deduped textured GLBs, and GALP outputs (layout_prediction.json + .glb) ready for Blender import. Accepts scene-dir (required), --gpu N (default 0), and --force (delete prior outputs + re-run everything). Trigger on "/stage1-initialize-scene".

Interior look-dev pass that enhances ONLY the lights and stage wall/floor/ceiling colors of an existing Blender `.blend` to match a reference image, leaving object materials and geometry untouched. Trigger on "/stage2-sub-env-enhance".

Build a separable stage (Floor, Wall_NN, Ceiling) inside a `.blend` by committing a rectilinear floor polygon in 2D from a pointmap + image, then extruding in 3D. Also vendors the layout_prediction.json → blender_scene.json → .blend converters (`src/convert.py`, `src/build.py`). Trigger on "/stage2-sub-pointmap-to-separable-stage".

Stage 3 orchestrator: scene-analyze-prepare (auto pre-step if needed) → auto-pass (no LLM) → planner-review (1 LLM call, revises operation plan + adds deletes) → apply revised plan → render → validation (1 LLM call) → island-refiner (needed groups only, N-iter Opus per group (default 20, configurable via num_max_iter) via stage3-sub-island-refiner) → merge-back → final render. Self-contained — auto-dispatches scene-analyze-prepare if its outputs are missing. Trigger on "/stage3-scene-refinement" or "/scene-refinement".

Per-group island refiner. Dispatched once per problem group (identified by validation) via the Task tool. Each dispatch drives the FULL iteration refinement loop (default 20 iters) inside a single `stage3-island-refiner` Opus subagent for ONE relation-group island (anchor + members). The agent refines the WHOLE GROUP PATTERN per iter — multi-member translation moves for spacing, and per-member yaw targets (NEVER uniform yaw deltas) for facing direction. Triggered indirectly via stage3-scene-refinement; do not invoke directly.

Given a scene_dir (image + .blend), extract three planning JSONs — `json/object_state.json`, `json/blend_info.json`, and `inputs/relation_graph.json` — without producing any operation plan or Blender edits. Trigger on "/scene-analyze-prepare".

Single-shot Stage 3 auto-pass — heuristic + graph_tool → operation_plan.json, apply, render. No LLM. Trigger ONLY on direct invocation ("/scene-refiner"), never auto-triggered from "fix the scene".

Stage-agnostic 5-view Cycles renderer (perspective / bev / wide / topcorner / topcorner_opposite) for any populated `.blend` — does not touch lights, world, materials, or camera. Usable from Stage 1 / Stage 2 / Stage 3 pipelines. Trigger on "/general-multi-view-render" or "just re-render the 5 views".

Full end-to-end pipeline orchestrator — runs stage1-initialize-scene → stage2-environment-construction → stage3-scene-refinement in sequence on a single scene_dir containing only image.png. Accepts: scene-dir (required), --gpu N (default 0, Stage 1 only), --island-refine-iter N (default 10, Stage 3 only), --force (skip all resume checks and rerun from Stage 1). Trigger on "/scene-orchestration", "run the full pipeline", "process scene end-to-end", "image to final 3D scene", or any time the user provides a scene_dir with only image.png and wants the complete output.

Render the SceneConductor pipeline-stage "progression" report videos (and a matching .blend per clip) from a processed scene_dir: object pop-in, Stage-2 environment build (interior + external), Stage-3 first refine (raw→planned), Stage-3 planned→final, and an isolated relation-group island refine (BEV + perspective). Use this whenever the user wants per-stage / per-step report videos, "stage progression" or "refinement process" clips, an env-build or wall-build video, a planned-vs-final or first-refine animation, or an island / table-and-chairs alignment clip from a SceneConductor scene — even if they only name one of the stages or say "make the report videos" or "render the stage clips like before". Produces 1160×600 H.264 mp4s on a white background with the established camera/crop/lighting conventions. NOT for the single 30s combined demo (that is scene-conductor-demo-video) and NOT for the 5-view look-dev render (general-multi-view-render).

Orchestrator that inspects a `scene_dir` and runs the Stage-2 sub-skills in order, skipping completed stages and stopping on blocked inputs. Trigger on "/stage2-environment-construction".