99 skills · sorted by relevance
Prepare, submit and reconcile reviewed Fusion Automation jobs with exact sources, scoped authorization and explicit budgets. Not for arbitrary model-generated cloud scripts.
The English summary is being prepared. Showing the original description.
Inspect an explicit Fusion design, parameters, geometry, references, configuration and health without editing it. Use for model understanding or engineering audits.
The English summary is being prepared. Showing the original description.
Expert-level nuclear physics knowledge. Use when working with nuclear structure, radioactive decay, nuclear reactions, fission, fusion, binding energy, nuclear models, radiation, or nuclear applications. Also use when the user mentions 'radioactive decay', 'half life', 'binding energy', 'fission', 'fusion', 'alpha decay', 'beta decay', 'gamma decay', 'nuclear reaction', 'cross section', 'Q value', 'nuclear reactor', or 'radiation dose'.
The English summary is being prepared. Showing the original description.
Generate images using AI models — Gemini, DALL·E, Stable Diffusion, Midjourney, Flux, Leonardo, Ideogram, and more. MCP bridge tools are primary; the Design Stack media worker is the fallback for advanced providers. Covers prompt engineering, model selection, size/aspect ratios, style parameters, batch generation, multi-site awareness, and WordPress integration. Use when you need to create hero images, social graphics, product mockups, brand illustrations, or any visual asset for marketing, social media, or e-commerce.
The English summary is being prepared. Showing the original description.
Build and publish a Gradio demo on Hugging Face Spaces for a user-provided LoRA. Use when someone asks to create, generate, ship, or publish a Space, demo, Gradio app, or playground for a LoRA — including LoRAs for Qwen-Image, Qwen-Image-Edit, LTX-Video, Wan, FLUX, SDXL, or other diffusion base models. Also triggers when someone describes a LoRA they trained or hosts on the Hub and wants to share it. Covers picking the right base pipeline and `diffusers` inference recipe, designing a UI tailored to the LoRA's task and inputs (Union/multi-task control, edit, video, image, etc.), respecting model-card recommendations (trigger words, steps, guidance, LoRA scale, example inputs), and shipping to ZeroGPU hardware as a private Space by default.
The English summary is being prepared. Showing the original description.
Expert en génération d'images ComfyUI (workflows, nodes, models, inpainting, upscaling)
The English summary is being prepared. Showing the original description.
Performs ML-based protein-ligand pose prediction and scoring using DiffDock-L (diffusion-based), Boltz-1 / Boltz-2 (foundation model with affinity), Chai-1, AlphaFold3 ligand, EquiBind, TANKBind, NeuralPLexer, and hybrid workflows (DiffDock pose + GNINA rescore + PoseBusters QC). Explicit handling of when ML beats classical docking, when classical beats ML, the PB-invalid pose problem, and rescoring as the standard production hybrid. Use when modern docking is needed: foundation-model ligand-pose prediction, AI rescoring of classical poses, or scaffold-hopping in cross-docking scenarios.
The English summary is being prepared. Showing the original description.
Performs ML-based protein-ligand pose prediction and scoring using DiffDock-L (diffusion-based), Boltz-1 / Boltz-2 (foundation model with affinity), Chai-1, AlphaFold3 ligand, EquiBind, TANKBind, NeuralPLexer, and hybrid workflows (DiffDock pose + GNINA rescore + PoseBusters QC). Explicit handling of when ML beats classical docking, when classical beats ML, the PB-invalid pose problem, and rescoring as the standard production hybrid. Use when modern docking is needed: foundation-model ligand-pose prediction, AI rescoring of classical poses, or scaffold-hopping in cross-docking scenarios.
The English summary is being prepared. Showing the original description.
Performs ML-based protein-ligand pose prediction and scoring using DiffDock-L (diffusion-based), Boltz-1 / Boltz-2 (foundation model with affinity), Chai-1, AlphaFold3 ligand, EquiBind, TANKBind, NeuralPLexer, and hybrid workflows (DiffDock pose + GNINA rescore + PoseBusters QC). Explicit handling of when ML beats classical docking, when classical beats ML, the PB-invalid pose problem, and rescoring as the standard production hybrid. Use when modern docking is needed: foundation-model ligand-pose prediction, AI rescoring of classical poses, or scaffold-hopping in cross-docking scenarios.
The English summary is being prepared. Showing the original description.
Advanced CV for infrastructure inspection including forest fire detection, wildfire precondition assessment, roof inspection, hail damage analysis, thermal imaging, and 3D Gaussian Splatting reconstruction. Expert in multi-modal detection, insurance risk modeling, and reinsurance data pipelines. Activate on "fire detection", "wildfire risk", "roof inspection", "hail damage", "thermal analysis", "Gaussian Splatting", "3DGS", "insurance inspection", "defensible space", "property assessment", "catastrophe modeling", "NDVI", "fuel load". NOT for general drone flight control, SLAM, path planning, or sensor fusion (use drone-cv-expert), GPU shader development (use metal-shader-expert), or generic object detection without inspection context (use clip-aware-embeddings).
The English summary is being prepared. Showing the original description.
Search arXiv for papers on a research topic and combine their methods, findings, and limitations into a structured literature review. Identify shared themes, conflicting results, and research gaps, with references formatted in APA, IEEE, or BibTeX.
Predict how a small molecule may sit within a protein using structural inputs. Generate ranked 3D poses without defining a search region, then use those poses as starting points for further evaluation.
Create figures for machine learning papers from text or experiment data. Uses Gemini for architecture diagrams and matplotlib/seaborn for numerical charts, choosing suitable layouts and chart types while preparing consistent styling, export files, and reusable generation scripts.
Designs complete research plans that integrate clinical variables with multi-omics data from a user-provided biomedical direction. Always use this skill whenever a user wants to design, scope, or structure a study that combines clinical variables with transcriptomics, proteomics, metabolomics, epigenomics, or related omics layers for mechanism interpretation, biomarker development, risk stratification, treatment-response analysis, or translational use. It should define the clinical use case, alignment across data layers, feature-reduction and fusion logic, modeling route, mechanism-interpretation layer, validation ladder, and four workload configurations (Lite / Standard / Advanced / Publication+). Never fabricate datasets, accession numbers, sample counts, metadata completeness, platform coverage, literature references, PMIDs, DOIs, or validation status. Always include the mandatory Dataset Disclaimer immediately before any workflow section that mentions datasets or public resources.
The English summary is being prepared. Showing the original description.
Profiles a genome from raw reads BEFORE assembly with a k-mer spectrum (KMC or Jellyfish histogram), then models it with GenomeScope2 to estimate genome size, heterozygosity, repeat content, and ploidy, and Smudgeplot to infer ploidy from heterozygous k-mer pairs (diploid AB vs triploid AAB vs tetraploid AABB). Covers choosing k via Merqury best_k.sh, the k-mer-coverage vs sequencing-coverage confusion, reading het/repeat/contamination/organelle peaks, why noisy ONT must not be used for counting, and how the estimate becomes the NG50 denominator, the Flye -g value, the hifiasm --hom-cov/purge setting, and the 1.5-2x-too-big haplotig sanity check. Use when starting any de novo assembly, deciding whether short reads can work, estimating genome size for an unknown organism, diagnosing ploidy, or sanity-checking an assembly's size against expectation.
The English summary is being prepared. Showing the original description.
Performs gene-level association from GWAS summary statistics via genetically predicted tissue expression using FUSION, PrediXcan, S-PrediXcan, S-MultiXcan, UTMOST, MOSTWAS, kTWAS, EpiXcan, TIGAR-V2, and probabilistic fine-mapping with FOCUS and MA-FOCUS. Use when running TWAS from GWAS sumstats, prioritising candidate causal genes from a GWAS lead locus, picking single-tissue vs cross-tissue models, identifying LD-induced TWAS false positives, choosing ancestry-matched prediction weights, fine-mapping co-regulated TWAS hits, or triangulating TWAS with cis-eQTL Mendelian randomization and colocalization to nominate a causal gene.
The English summary is being prepared. Showing the original description.
Use when generating or editing images with Google's Gemini image models (Nano Banana) on Scenario via MCP: text-to-image, natural-language instruction editing, identity locking or style transfer from reference images, multi-image fusion, pulling stills from a video clip, Google Search grounding, thinking level tuning, 512 to 4K output, or choosing between Flash, Pro, and Lite. Keywords: Gemini 3.1 Flash, Gemini 3.0 Pro, Gemini 3.1 Lite, Nano Banana 2, Nano Banana Pro, txt2img, img2img.
The English summary is being prepared. Showing the original description.
Generate AI images using Volcengine Seedream model. Supports text-to-image (T2I), image editing (I2I), multi-image fusion, and web-search-based generation. Use this skill when the user wants to create, generate, or edit images.
The English summary is being prepared. Showing the original description.
Molecular docking prediction using DiffDock diffusion model
The English summary is being prepared. Showing the original description.
Provides guidance for writing and benchmarking optimized CUDA kernels for NVIDIA GPUs (H100, A100, T4) targeting HuggingFace diffusers and transformers libraries. Kernels must be kernel-builder/ABI3-compliant: no pybind11, no setup.py, TORCH_LIBRARY_EXPAND bindings only. Supports models like LTX-Video, Stable Diffusion, LLaMA, Mistral, and Qwen. Includes integration with HuggingFace Kernels Hub (get_kernel) for loading pre-compiled kernels. Includes benchmarking scripts to compare kernel performance against baseline implementations.
The English summary is being prepared. Showing the original description.
Routes official FlashVSR diffusion-based streaming video super-resolution setup, model preparation, and CUDA inference for v1 and v1.1.
The English summary is being prepared. Showing the original description.
Use and troubleshoot the Dream Textures Blender add-on for Stable Diffusion image generation, texture projection, render passes, model setup, and backend APIs.
The English summary is being prepared. Showing the original description.
Use the DALLE2-pytorch package for DALL-E 2 style text-to-image model APIs, diffusion prior/decoder training, WebDataset data layouts, and tracker/checkpoint workflows.
The English summary is being prepared. Showing the original description.
Use when coordinating complex tasks across multiple AI agents with a centralized handoff document for planning, execution tracking, and feedback fusion.
The English summary is being prepared. Showing the original description.