Interactive first-time-user walkthrough for the Video Ingestion Agent (segmentation -> entity graph DB -> retrieval -> webapp). Drives the user through setup ONE STEP AT A TIME with verification between steps, not as a reference dump. Use this skill whenever…
Troubleshooting skill for robotic_grounding failures — Docker/container issues, GPU-not-visible, missing MANO or object assets, motion-file-not-found, Isaac Lab startup crashes, permission errors, and pipeline-stage failures. Use this skill whenever something…
Interactive first-run walkthrough for robotic_grounding (build the Docker image → provide MANO + a dataset → run the retarget pipeline → dummy-agent smoke test → optional train/eval). Drives the user ONE STEP AT A TIME with verification between steps, not as…
Guide for generating robot-behaviour datasets from trained Sharpa (floating-hand) V2D policies — rolling out checkpoints in Isaac Lab with cameras and exporting to LeRobot. Use this skill whenever the user wants to generate/record a dataset from a trained…
Launch, monitor, resume, and verify this repository's BundleSDF or SAM3D HOI object reconstruction pipeline. Use when a user asks to run object reconstruction, process calibrated stereo mapping data, try the included basketball example, produce a BundleSDF…
Diagnose and repair failures in this repository's mesh-to-USD generation, scoped USD validation, recorded-support inference, FoundationPose calibration, catalog batching, or Isaac Sim drop tests. Use when Docker, GPU, EULA, weights, input contracts, CoACD…
Generate, validate, drop-test, resume, and batch-process rigid USD assets from reconstructed object meshes or exported HOI sequences in this repository. Use when a user asks to convert a GLB to USD, create a physics-enabled rigid object, infer a recorded…
Prepare this repository's mesh-to-USD generation, structural validation, and Isaac Sim drop-test environment. Use when a user asks to build the mesh-to-USD images, verify Docker/GPU access, prepare FoundationPose support calibration, check model weights,…
Interactive walkthrough for turning a NEW ego-video reconstruction (a result.npz bundle) into a trainable robotic_grounding sequence — validate the bundle, level it, then retarget to the floating Sharpa hand and/or the Dexmate Vega whole body, build object…
Run and validate the repository-local multi-view camera calibration and human-object reconstruction pipelines. Use when Codex is asked to prepare, launch, monitor, troubleshoot, or verify `v2d.pipelines.run_mv_calibration` or…
Prepare this repository’s egocentric reconstruction pipeline. Use when a user asks to install host packages, build Docker images, download model weights, verify Docker/GPU prerequisites, or diagnose missing setup for…
Diagnose and repair failures in this repository's BundleSDF or SAM3D HOI object reconstruction workflow. Use when setup checks fail, Docker or GPU access is broken, weights or images are missing, input preflight fails, CuSFM or scan-quality gates fail, a…
Prepare this repository's HOI object reconstruction environment for BundleSDF or SAM3D. Use when a user asks to install host orchestration packages, validate Docker/GPU access, build missing reconstruction images, download mode-specific weights, preflight…
Generates the exact command to run the robotic_grounding floating-hand (Sharpa Wave / Dex3) pipeline — retargeting, dummy-agent smoke tests, RL training, and evaluation. Use this skill whenever the user wants to RUN something in robotic_grounding and needs…
Pointer skill for the robotic_grounding G1 WHOLE-BODY humanoid pipeline (SONIC controller + RL residuals). Orients the user on the two reference-motion sources and the SonicG1 Isaac Lab tasks, then points at the authoritative source READMEs for exact…
Run this repository’s egocentric reconstruction pipeline on a video. Use when a user asks to process an MP4 with HaMeR or DynHaMR hand tracking, an object prompt or mesh, undistortion, DROID-SLAM, gravity alignment, gsplat refinement, or a Three.js result…
Prepare the repository-local egocentric reconstruction pipeline for Codex-driven work. Use when a user asks to finish setup, install host orchestration packages, build Docker images, download model weights, check GPU/Docker prerequisites, or debug missing…
Run the egocentric reconstruction pipeline on an input video. Use when a user asks Codex to process an MP4 or video with the reconstruction ego pipeline, choose between HaMeR/WiLoR and DynHaMR hand tracking, pass an object prompt or object mesh, enable…
Diagnostic checklist for the Video Ingestion Agent environment — run when something is broken, before a first ingest, or as a sanity check. Use this skill whenever the user reports the pipeline isn't working, the vLLM server won't start or crashes during…