592 skills · sorted by relevance
Expert blueprint for AI pathfinding (tower defense, RTS, stealth) using NavigationAgent2D/3D, NavigationServer, avoidance, and dynamic navigation mesh generation. Use when implementing enemy AI, NPC movement, or obstacle avoidance. Keywords NavigationAgent2D, NavigationRegion2D, pathfinding, NavigationServer, avoidance, baking, NavigationObstacle.
The English summary is being prepared. Showing the original description.
AiZynthFinder retrosynthetic route planning (CASP) from AstraZeneca Molecular AI. Monte Carlo tree search guided by a template-based neural expansion policy recursively disconnects a target SMILES until precursors are found in a purchasable stock. Covers config.yml (v4 format), aizynthcli batch screening, the AiZynthFinder/AiZynthExpander Python API, one-step disconnections, custom stocks via smiles2stock, scorers, Retro*/breadth-first/DFPN search alternatives, and reading output.json.gz / trees.json. Use for synthesis route planning, synthesizability screening, and building-block/precursor search. For reaction barriers use neb-irc-activation-energy; for 2D reaction scheme drawing use rdkit-chemdraw-cdxml.
The English summary is being prepared. Showing the original description.
Expert in designing and implementing intelligent game AI systems including behavior trees, finite state machines, GOAP, utility AI, pathfinding, steering behaviors, and perception systems. Specializes in creating believable, performant NPC behaviors that enhance player experience. Use when "game AI, NPC behavior, behavior tree, state machine for game, enemy AI, pathfinding, A* algorithm, navmesh, steering behavior, GOAP, utility AI, AI perception, combat AI, companion AI, boss AI, crowd simulation, flocking, game-ai, behavior-trees, pathfinding, npc, state-machines, goap, utility-ai, steering, perception" mentioned.
The English summary is being prepared. Showing the original description.
"Implement VCG mechanism for incentive-compatible ad slot allocation with truthful bidding. Use this skill when the user needs to design a truthful auction mechanism, compute externality-based payments, or understand why platforms may prefer GSP over VCG — even if they say 'truthful auction design', 'VCG payments', or 'incentive-compatible mechanism'.".
The English summary is being prepared. Showing the original description.
Prepare high-quality datasets for LLM fine-tuning with filtering, deduplication, augmentation, and RLHF data formatting. Activate on: fine-tuning data, training data curation, RLHF dataset, data quality filtering, SFT dataset. NOT for: model training infrastructure (ai-engineer), prompt engineering without fine-tuning (prompt-engineer).
The English summary is being prepared. Showing the original description.
Knowledge base from the FAA Human Factors Design Standard (HF-STD-001B, 2016) — the FAA's consolidated human-factors / human-systems-integration design criteria for systems it manages, operates, or maintains. Use for requirement-level design rules on: general human-factors design principles; automation and function allocation; designing equipment for maintenance; displays, controls, and visual indicators; alarms, audio, and voice communications; the computer-human interface (information presentation, coding, interaction styles, windows, dialogue); keyboards, input devices, and workstation/workplace ergonomics; system security, personnel safety, environment, anthropometry, and user documentation; plus intended-use and tailoring guidance. It is a quantified, source-attributed synthesis (heavily citing MIL-STD-1472G, MIL-HDBK-759C, DOD-HFDG-ATCCS, NUREG-0700, NASA-STD-3000, ANSI, OSHA, ISO 9241). Complements nasa-hsi as a parallel HSI tradition from the civil-aviation domain. Scope limits: it is design criteria, NOT an HSI lifecycle/process model (use nasa-hsi, nasa-npr-7123, or dau-se-guidebook for process); it does NOT reproduce the third-party standards it cites; it is FAA-acquisition-oriented (ATC, towers, TRACONs, airway facilities), not a general consumer-UX guide; and it is thin on modern touchscreen/mobile, web, and AI-driven interface patterns and on quantitative human-reliability analysis.
The English summary is being prepared. Showing the original description.
AiZynthFinder retrosynthetic route planning (CASP) from AstraZeneca Molecular AI. Monte Carlo tree search guided by a template-based neural expansion policy recursively disconnects a target SMILES until precursors are found in a purchasable stock. Covers config.yml (v4 format), aizynthcli batch screening, the AiZynthFinder/AiZynthExpander Python API, one-step disconnections, custom stocks via smiles2stock, scorers, Retro*/breadth-first/DFPN search alternatives, and reading output.json.gz / trees.json. Use for synthesis route planning, synthesizability screening, and building-block/precursor search. For reaction barriers use neb-irc-activation-energy; for 2D reaction scheme drawing use rdkit-chemdraw-cdxml.
The English summary is being prepared. Showing the original description.
Use Hugging Face Hub via MCP server tools. Search models, datasets, Spaces, papers. Get repo details, fetch documentation, run compute jobs, and use Gradio Spaces as AI tools. Available when connected to the HF MCP server.
The English summary is being prepared. Showing the original description.
Pick the right serving container for a SageMaker model deployment and find its current image URI. Use this skill whenever about to deploy a model to a SageMaker endpoint and an image URI needs to be chosen — including when the user says "deploy this LLM", "host this HuggingFace model", "serve this fine-tuned model", "deploy this embedding model", "host a reranker", "serve a sentence-transformers model", or when about to hardcode any container URI in deployment code. HuggingFace-curated Deep Learning Containers are ALWAYS preferred: HuggingFace vLLM (LLMs and generative rerankers), HuggingFace vLLM-Omni (multimodal), TEI (embeddings/cross-encoder rerankers), HF Inference Toolkit (other transformers). Generic images (AWS vLLM, DJL-LMI, SGLang) are used only when no HuggingFace image is compatible — never merely because they carry a newer version. Never hardcode a container URI from memory and never default to TGI. Prevents stale-image failures and wrong-region URIs.
The English summary is being prepared. Showing the original description.
Set up and run the full RLHF pipeline (SFT, reward model training, RL from reward model) using the Tinker API. Use when the user wants to do RLHF, train a reward model, or run the full preference-based RL pipeline.
The English summary is being prepared. Showing the original description.
Execute model inference on GPU cloud providers. Handles code generation, deployment, execution, and result collection across HF Inference API/Endpoints, Colab, Modal, beam.cloud, Vast.ai, and RunPod. Use when running models on GPU, deploying to cloud, executing notebooks, or troubleshooting GPU execution failures. Triggers on "run on GPU", "execute model", "deploy to modal", "colab notebook", "beam deploy", "HF inference", "HF endpoints", "vast", "runpod".
The English summary is being prepared. Showing the original description.
Prepare high-quality datasets for LLM fine-tuning with filtering, deduplication, augmentation, and RLHF data formatting. Activate on: fine-tuning data, training data curation, RLHF dataset, data quality filtering, SFT dataset. NOT for: model training infrastructure (ai-engineer), prompt engineering without fine-tuning (prompt-engineer).
The English summary is being prepared. Showing the original description.
Map a codebase into feature flowcharts, find duplicated logic, and propose a simpler shared architecture. Produce source-linked diagrams, an evidence-based duplication report, and prompts for handing each proposed system to /make-plan.
Design rewards for reinforcement learning, where feedback guides model behavior. Choose how to score outcomes and intermediate actions, check for scoring loopholes, and use independent tests to detect incomplete work presented as finished.
Tailor a resume to a specific job by revising its summary, reordering skills and achievements, and incorporating accurate job keywords. Document the changes and maintain targeted versions while keeping experience, qualifications, and metrics truthful.
Assess language models with over 60 academic benchmark tests using consistent prompts and scoring. Compare models, track performance during training, and organize research results with support for HuggingFace models, vLLM, and APIs.
Assess language models with over 60 academic benchmark tests using consistent prompts and scoring. Compare models, track performance during training, and organize research results with support for HuggingFace models, vLLM, and APIs.
Fine-tune language models with TRL using GRPO, a method that compares scored responses to the same prompt. Covers training data, scoring functions, GPU-conscious settings, and monitoring for tasks with measurable correctness or required output formats.
Choose a serving container for models deployed on SageMaker. Match the model type to a compatible image, retrieve its address from the AWS catalog, and check regional availability, CPU or GPU variants, and required host settings.
Evaluate Hugging Face Hub models using inspect-ai or lighteval, with local inference or a provider-backed option. Choose an inference engine, start with a small test, and switch engines when model compatibility requires it.
Build and maintain machine learning apps on Hugging Face Spaces. Choose a suitable app framework and computing hardware, load models, deploy files, and check logs and outputs to diagnose build or runtime problems.
Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing repositories, models, datasets, and Spaces on the Hugging Face Hub. Replaces now deprecated `huggingface-cli` command.
The English summary is being prepared. Showing the original description.
Add AI navigation in Unity 6.3 LTS: bake a NavMesh with the AI Navigation package (NavMeshSurface), move agents with NavMeshAgent.SetDestination, and handle dynamic obstacles. Use when setting up pathfinding, making an enemy chase the player, baking navigation, or when the user mentions NavMesh, NavMeshAgent, NavMeshSurface, NavMeshObstacle, or Unity pathfinding.
The English summary is being prepared. Showing the original description.
Design NPC and enemy decision-making with finite state machines, behavior trees, steering behaviors, and A* pathfinding — engine-neutral algorithms that pair with the detected engine's navigation API. Use when building enemy AI, an FSM or behavior tree, steering/flocking, or pathfinding, or when the user mentions state machine, behavior tree, blackboard, A*, navmesh, seek, or patrol/chase.
The English summary is being prepared. Showing the original description.